Datasets:

Interhuman
/

wave1-test-set

Agreement: Nods, verbal affirmations, affiliative smiles, posture mirroring
Disagreement: Opposition, interruptions, head shakes, eye rolls
Engagement: Sustained orientation, eye contact, forward-leaning posture
Disengagement: Gaze aversion, fidgeting, low-energy speech
Confusion: Puzzled facial display, increased blinking, clarification questions
Hesitation: Long gaps before speaking, filled pauses, restarts
Uncertainty: Hedged answers, rising final contour, shoulder shrugs
Skepticism: Questioning stance, delayed responses, furrowed brows
Confidence: Upright posture, sustained eye contact, clear speech
Frustration: Tension cues, higher pitch, sighs, complaints
Interest: Attentional engagement, widened eyes, faster speech
Stress: Self-touch, repetitive fidgeting, rapid blinking, vocal instability

Data Format

The dataset is provided as a JSONL file where each line contains:

{
  "messages": [
    {
      "content": "System prompt with task instructions...",
      "role": "system"
    },
    {
      "content": "<video>",
      "role": "user"
    },
    {
      "content": "",
      "role": "assistant"
    }
  ],
  "videos": ["https://..."],
  "multilabels": ["Signal1", "Signal2", ...],
  "video_id": "unique_video_identifier.mp4"
}

Usage

from datasets import load_dataset

# Load dataset
dataset = load_dataset("Interhuman/wave1-test-set", split="train")

# Or load the raw JSONL file
import json

with open("configs_wave1-test-set-dataset.jsonl", "r") as f:
    data = [json.loads(line) for line in f]

# Access first sample
sample = data[0]
print(sample['video_id'])
print(sample['multilabels'])
print(sample['videos'][0])

Dataset Statistics

Total Samples: 50
Multi-label Classification: Each video can have multiple social signals
Signal Distribution: Varies across samples

Citation

If you use this dataset, please cite:

@dataset{interhuman_wave1_test,
  title={Wave1 Test Set - Social Signal Detection},
  author={Interhuman},
  year={2024},
  publisher={Hugging Face}
}

License

Apache 2.0

Downloads last month: -

Size of downloaded dataset files:

664 kB

Size of the auto-converted Parquet files:

102 kB

Number of rows:

146

messages listlengths 3 3	videos listlengths 1 1	multilabels listlengths 0 4	video_id stringlengths 41 149
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Adam%20Rosser%20Interviews%20Duncan%20Jones--0f9fpwGsyg_seg_00010_shots_3.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement", "Interest" ]	Adam Rosser Interviews Duncan Jones--0f9fpwGsyg_seg_00010_shots_3.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Adam%20Rosser%20Interviews%20Duncan%20Jones--0f9fpwGsyg_seg_00010_shots_5.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=d57c69af02ad55359209276bb5d3b1405f450d4903a7021b0c1ff4fc20478446ad5be49c2d81b085c9195206b821d46da6c9c39b0537bbe9a2dd1d9ad62586d9dcc107cbd99ad136592209c0c1c5717a9e1d9dab944bf717297ccb313eb625e5f0efc5fca1ed05563142db3d2d7463b676b60c1c170c9249e101921ef4a7f25accf94b08a1b9c8e8da5915dce86b27df6dd80e5017ae19df8899f51a9f452ab11ec8695ed174b0fada221c1d9774cdf35cd82a676b539ebc9b6714342a5f53a8895fccd590c2380edda7a28f7b130463ff746fd07c817b5e1c65bcc81c7d389be88f7b4f458995e8523415389b34e557316e6d4c4d8b318ba8063314d0c2efce" ]	[ "Confidence" ]	Adam Rosser Interviews Duncan Jones--0f9fpwGsyg_seg_00010_shots_5.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/ECB%20Public%20Hearing%20on%20regulation%20and%20guide%20on%20options%20and%20discretions-7ZscTxWqj4c_seg_00018_shots_0.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=abb5f5c7707a77e599c0d81d1a7842eb49854faf3eb72b3002d92fcbb030430c66c26a2fcd29f4e93b5294507645dcdca14c707c65d49126db38a146783c34c78153b596f773b18b2e60552c4e4b5ee730e8a1eaec64481e17a7b4e03eb67676fb639e56618661fa4bdebbaa7b2c7bd93f6fab76207017ab11b3447db920441586bdc6e6a529ad226b91649994c6e2893febe52115f5cf6588aefb2156a18f4fa203df7b09a73bb5b79ca4d6cd176fcf2e18d0dec2ffe241b0a79e7461ac6e7d111698bb49010753ad19a31b80779020db1ce2361b556119fb9361412ca91c31ec28d5c231e16d066f3d8492e03056569d8ebbd0a739c960727b176e69db1b75" ]	[ "Disengagement", "Skepticism" ]	ECB Public Hearing on regulation and guide on options and discretions-7ZscTxWqj4c_seg_00018_shots_0.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/ECB%20Public%20Hearing%20on%20regulation%20and%20guide%20on%20options%20and%20discretions-7ZscTxWqj4c_seg_00026_shots_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	ECB Public Hearing on regulation and guide on options and discretions-7ZscTxWqj4c_seg_00026_shots_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Fix%20Pre-Interview%20Anxiety%20%28step%20by%20step%29-q4NFCFw2rgU_seg_00008_shots_0.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	Fix Pre-Interview Anxiety (step by step)-q4NFCFw2rgU_seg_00008_shots_0.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Fix%20Pre-Interview%20Anxiety%20%28step%20by%20step%29-q4NFCFw2rgU_seg_00008_shots_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	Fix Pre-Interview Anxiety (step by step)-q4NFCFw2rgU_seg_00008_shots_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Fix%20Pre-Interview%20Anxiety%20%28step%20by%20step%29-q4NFCFw2rgU_seg_00008_shots_3.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	Fix Pre-Interview Anxiety (step by step)-q4NFCFw2rgU_seg_00008_shots_3.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Fix%20Pre-Interview%20Anxiety%20%28step%20by%20step%29-q4NFCFw2rgU_seg_00008_shots_5.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	Fix Pre-Interview Anxiety (step by step)-q4NFCFw2rgU_seg_00008_shots_5.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/How%20To%20Control%20Your%20Anger%EF%BC%9A%20Accepting%20Your%20Emotions%20%23shorts-AbREVYbHBiM_seg_00002_shots_5.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122102Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement", "Uncertainty" ]	How To Control Your Anger： Accepting Your Emotions #shorts-AbREVYbHBiM_seg_00002_shots_5.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Mayor%20Eric%20Adams%20Holds%20Public%20Hearing-KjXWeHg-kG8_seg_00012_shots_3.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	Mayor Eric Adams Holds Public Hearing-KjXWeHg-kG8_seg_00012_shots_3.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Mayor%20Eric%20Adams%20Holds%20Public%20Hearing-KjXWeHg-kG8_seg_00023_shots_1.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence" ]	Mayor Eric Adams Holds Public Hearing-KjXWeHg-kG8_seg_00023_shots_1.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Mayor%20Eric%20Adams%20Holds%20Public%20Hearing-KjXWeHg-kG8_seg_00023_shots_7.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence" ]	Mayor Eric Adams Holds Public Hearing-KjXWeHg-kG8_seg_00023_shots_7.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Troubled%20by%20Fear%EF%BC%9F%20Just%20Change%20Your%20Channel%21%20-%20Sadhguru-K7hYf0yIK5w_seg_00004_shots_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence" ]	Troubled by Fear？ Just Change Your Channel! - Sadhguru-K7hYf0yIK5w_seg_00004_shots_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Troubled%20by%20Fear%EF%BC%9F%20Just%20Change%20Your%20Channel%21%20-%20Sadhguru-K7hYf0yIK5w_seg_00005_shots_5.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	Troubled by Fear？ Just Change Your Channel! - Sadhguru-K7hYf0yIK5w_seg_00005_shots_5.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Troubled%20by%20Fear%EF%BC%9F%20Just%20Change%20Your%20Channel%21%20-%20Sadhguru-K7hYf0yIK5w_seg_00005_shots_6.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=cc2305ce0ba07d2e357fdb6924c7b6c91fdb540296440c7a39999874098f93e52a4fd7d41a37a57b4d9dd6d94ac6956f89f6b7db4c944902d530a34276bcff783f641cc5a99c8288bebf3ec66a414cdf13887c6e48a0a170995f1c8ed281b3fdb4d759fe85fececad6dddd36916fa1c1d1cadfce3a6cd10e25e2cbcf9f09eb0c3fcfd03fcd35bf1327b56f46982e5ed1510a305209d38414c4b1f9e6d4155d9a56c4b6b105d077e68c4add715f7a3d0fad71f213b2310080f5f68aabec0303df31b188b8ac000cb7870302a110fdff9d749a3062e9c637589cb171a4f8bda0c660969c0b1fc5a4d169a0f34f527d29497933af828b1a7f93dedb6602e8beb236" ]	[ "Confidence", "Engagement" ]	Troubled by Fear？ Just Change Your Channel! - Sadhguru-K7hYf0yIK5w_seg_00005_shots_6.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Whitney%20Houston%E2%80%99s%20response%20to%20being%20vocally%20compared%20to%20Mariah%20Carey%20in%201991.-iing0zpaY6E_seg_00002_shots_0.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Hesitation", "Uncertainty" ]	Whitney Houston’s response to being vocally compared to Mariah Carey in 1991.-iing0zpaY6E_seg_00002_shots_0.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/Whitney%20Houston%E2%80%99s%20response%20to%20being%20vocally%20compared%20to%20Mariah%20Carey%20in%201991.-iing0zpaY6E_seg_00002_shots_4.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=cce9dc2fb8336386f49924b9dfa7263f8f38fa913815ca6eaa14d9d91e07bafcafa733b80483eed21b254f7b3c9d495aac1e59878e74e2166fd46bdbc2ca29f644a47903c20f3393a14129700a3eb7b8794dcebd0d87b1e5472e10dffbd0f7d07ca88f6aaf28b907781cd1009c6e1cd7ca3aa021e1cb39b139d4ec749b7c28c700df762dcfc838260eab926df19b80ca1761e2ece03ca7b6ff5d27bb33f40b183ae3e513704d68f3190a72be018a309cb302f996050c707658df14b54e435b44c6ed8e4240a47eb91ec99a04cc28e03e77f27eabe0235b05a4734441061e4222ca78d2ee9598d09a324299507465cef4c673b4e6879aad976d549fc0ac36e444" ]	[ "Confidence", "Disagreement", "Uncertainty" ]	Whitney Houston’s response to being vocally compared to Mariah Carey in 1991.-iing0zpaY6E_seg_00002_shots_4.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-5Y6mKSlHwTE_seg_00036_shots_0.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement", "Hesitation" ]	wave1-1-5Y6mKSlHwTE_seg_00036_shots_0.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-5Y6mKSlHwTE_seg_00036_shots_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement", "Hesitation", "Uncertainty" ]	wave1-1-5Y6mKSlHwTE_seg_00036_shots_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-5Y6mKSlHwTE_seg_00036_shots_4.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement", "Hesitation", "Interest" ]	wave1-1-5Y6mKSlHwTE_seg_00036_shots_4.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-5w6huC55Qas_seg_00001_shots_7.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	wave1-1-5w6huC55Qas_seg_00001_shots_7.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-5w6huC55Qas_seg_00038_shots_6.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	wave1-1-5w6huC55Qas_seg_00038_shots_6.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-5w6huC55Qas_seg_00086_shots_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	wave1-1-5w6huC55Qas_seg_00086_shots_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-5w6huC55Qas_seg_00086_shots_6.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement", "Interest" ]	wave1-1-5w6huC55Qas_seg_00086_shots_6.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-A%20Conversation%20With%20Stephen%20King%20%281_38_42%29-NLfINvRut4I_seg_013_7.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement" ]	wave1-1-A Conversation With Stephen King (1_38_42)-NLfINvRut4I_seg_013_7.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Anger%20problems.%20Is%20anger%20the%20real%20problem%EF%BC%9F-s2MB9JUAbRU_seg_006_1.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confusion", "Hesitation", "Uncertainty" ]	wave1-1-Anger problems. Is anger the real problem？-s2MB9JUAbRU_seg_006_1.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Anger%20problems.%20Is%20anger%20the%20real%20problem%EF%BC%9F-s2MB9JUAbRU_seg_008_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Disengagement", "Frustration", "Hesitation" ]	wave1-1-Anger problems. Is anger the real problem？-s2MB9JUAbRU_seg_008_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Dave%20Smith%20and%20Alex%20Nowrasteh%20debate%20immigration-6xWiSBnwIQM_chopped_seg_00034_shots_1.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Disagreement", "Engagement" ]	wave1-1-Dave Smith and Alex Nowrasteh debate immigration-6xWiSBnwIQM_chopped_seg_00034_shots_1.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Dave%20Smith%20and%20Alex%20Nowrasteh%20debate%20immigration-6xWiSBnwIQM_chopped_seg_00034_shots_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=d87765d0ac7f187d379bbc7d5e15d0c7b36e2302886d57be314f079a433540db46b4216c8cbef476d9ac2fe02054425da2c67b1a21a1cda9d3b7af4eb10107308abb6bf8cc22758ac1bebc69a0e6ed457c8a6500b49baf7ac87cf48f165b4fe5dbe68dca26ab5e9fd64b29924dc2f471042dfe3782fd0eb09b9f6b525c7e9ce761ba40116520d6f2aed0a7a59c14a84ebbe2936df538accc2e5b7b762dd9862f0fe303fa0fa5b1241893fef3dccad38362674413a9f83e74e98ad989463b18060ae9dbd8a9883c3cfe1212a6e0cfc98f5796334d34b0cf0119515c8028a39b52e9927e87fc0ea0a199fb8dedb41ecf16f4843a8f49bdcfd19b5689961dae72f7" ]	[ "Engagement", "Frustration" ]	wave1-1-Dave Smith and Alex Nowrasteh debate immigration-6xWiSBnwIQM_chopped_seg_00034_shots_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Dave%20Smith%20and%20Alex%20Nowrasteh%20debate%20immigration-6xWiSBnwIQM_chopped_seg_00034_shots_4.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement", "Frustration" ]	wave1-1-Dave Smith and Alex Nowrasteh debate immigration-6xWiSBnwIQM_chopped_seg_00034_shots_4.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Dave%20Smith%20and%20Alex%20Nowrasteh%20debate%20immigration-6xWiSBnwIQM_chopped_seg_00065_shots_0.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Disagreement", "Engagement", "Frustration", "Skepticism" ]	wave1-1-Dave Smith and Alex Nowrasteh debate immigration-6xWiSBnwIQM_chopped_seg_00065_shots_0.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Dave%20Smith%20and%20Alex%20Nowrasteh%20debate%20immigration-6xWiSBnwIQM_chopped_seg_00065_shots_1.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Disagreement", "Engagement", "Skepticism" ]	wave1-1-Dave Smith and Alex Nowrasteh debate immigration-6xWiSBnwIQM_chopped_seg_00065_shots_1.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Dave%20Smith%20and%20Alex%20Nowrasteh%20debate%20immigration-6xWiSBnwIQM_chopped_seg_00065_shots_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement", "Skepticism" ]	wave1-1-Dave Smith and Alex Nowrasteh debate immigration-6xWiSBnwIQM_chopped_seg_00065_shots_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Dave%20Smith%20and%20Alex%20Nowrasteh%20debate%20immigration-6xWiSBnwIQM_chopped_seg_00065_shots_5.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement" ]	wave1-1-Dave Smith and Alex Nowrasteh debate immigration-6xWiSBnwIQM_chopped_seg_00065_shots_5.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-GfejeUhU-Co_seg_00005_shots_0.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confidence", "Engagement" ]	wave1-1-GfejeUhU-Co_seg_00005_shots_0.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-GfejeUhU-Co_seg_00005_shots_6.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=1d4260cb7cee3f8083379aa00d5c5044d1e4c1c156c84c43751c67fdd829f5d18a2965c68cb80b56e79b0be200294bf8f140876c9e2314d86c1988092041d3729d31d4cec00090e5eae36e6970ac6f7efe33961f6bbf7d6bc1704787c0c06165631b1dacc9341779a1730468f74e63de83d15059d66b9c2b153dd3d1845937af407ba408f2823861b1b2fb0542fd991f4d92f38fb9eb31ee4920035772fe4e21a3e4dbb67c23c51d0f1d19f982c733cfb165de803f0a08d2295671f6d92fdc887cb0dfee13cd23e4f35cf1943c2747e083ac6eb2e87612d7dfbd64d5d5cf2d7673a5ae42ff0e025e22ebbb50338622d4676814720b5c4d05a23bef65c8c8d37d" ]	[ "Confidence" ]	wave1-1-GfejeUhU-Co_seg_00005_shots_6.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-GfejeUhU-Co_seg_00069_shots_2.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Hesitation" ]	wave1-1-GfejeUhU-Co_seg_00069_shots_2.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-GfejeUhU-Co_seg_00069_shots_4.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Confusion", "Frustration", "Hesitation", "Uncertainty" ]	wave1-1-GfejeUhU-Co_seg_00069_shots_4.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-Have%20to%20Make%20the%20Plays%20When%20They%20Come%20%EF%BD%9C%20Dylan%20Cole%20Player%20Interview-8AWpW2qLgeI_seg_002_4.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement", "Interest" ]	wave1-1-Have to Make the Plays When They Come ｜ Dylan Cole Player Interview-8AWpW2qLgeI_seg_002_4.mp4
[ { "content": "You are an expert behavioral analyst specializing in the real-time detection, classification, and annotation of social and affective signals from multimodal data streams (video, audio, and text transcripts if available).\n\n## CORE MANDATE\n* This is a multi-label, multi-class classification problem, your task is to detect the presence of the below listed social signals in the video material.\n* Analyze the provided video (incl. audio signal and video frames) and identify any occurrences of the following 12 social signals.\n* You must use the definitions and key behavioral manifestations provided below as your absolute criteria.\n* Only detect a social signal if there is good to high confidence.\n* It is allowed to detect multiple signals, or None.\n* Only return the detectec signals\n\n## SOCIAL SIGNAL DEFINITIONS\n\| Signal Name \| Key Behavioral Manifestations \|\n\| ----- \| ----- \|\n\| Agreement \| Nods timed to speaker phrasing, immediate verbal affirmations (\"yes,\" \"yeah\"), affiliative smiles, posture/gesture mirroring, brief mutual gaze synchrony. \|\n\| Disagreement \| Open opposition (challenge-format turns, rejections), interruptions, rejection displays (head shake/roll, eye roll), orienting-away/blocking actions (arm folding), stance-prefacing vocal cues (sighs, throat-clearing). \|\n\| Engagement \| Sustained orientation (on-target gaze, upright/forward-leaning posture), timely nods, short verbal backchannels (\"mm-hm\"), body facing interaction, minimal off-task activity. \|\n\| Disengagement \| Orienting-away (gaze aversion, head tilt), low-arousal signals (fidgeting, yawning, audible sighs), low-energy speech, minimal affiliation, very brief answers. \|\n\| Confusion \| Puzzled facial display (furrowed brow, tightened lids), increased blinking, brief gaze away from materials, silence beyond expected window, clarification questions targeting terms or steps. \|\n\| Hesitation \| Longer-than-usual gap before speaking, filled pauses (\"uh/um\"), restarts/self-repair, delayed/mitigated responses, final-syllable lengthening, creaky voice, brief gaze aversion. \|\n\| Uncertainty \| Low commitment (hedged answers, \"I think,\" \"I guess\"), long response gaps, rising final contour, visual cues (small shoulder shrug, palm-up hands, mouth-shrug lip stretch). \|\n\| Skepticism \| Questioning/doubting stance, delayed/mitigated responses (\"well... I'm not sure\"), prosodic tension (higher pitch), visible rejections (head shakes, furrowed brows), frequent follow-up probes. \|\n\| Confidence \| Upright open posture, sustained eye contact, audible speech with few disfluencies, wording avoiding hedges, neutral/serious facial set. \|\n\| Frustration \| Tension cues (firm lip press, upper-lip raiser, chin raiser, self-touch), vocal signs (higher pitch, louder intensity, sighs), linguistic markers (complaints, repeated negation). \|\n\| Interest \| Sustained attentional engagement (forward-leaning posture, head tilts), widened/tightened eyes, parted lips or raised chin, faster speech or greater vocal range, backchannels that invite continuation. \|\n\| Stress \| Heightened arousal/self-regulatory strain (self-touch, repetitive fidgeting, rapid blinking, sweating), vocal signs (higher pitch, instability), faster/shorter responses, filled pauses, simplified language. \|", "role": "system" }, { "content": "<video>", "role": "user" }, { "content": "", "role": "assistant" } ]	[ "https://storage.googleapis.com/ih-ground-truth-wave1/videos/wave1-1-LIVE%20NURSING%20INTERVIEW-Aj0nf3rKG_E_seg_003_1.mp4?X-Goog-Algorithm=GOOG4-RSA-SHA256&X-Goog-Credential=vlm-training%40interhuman-ai-prod.iam.gserviceaccount.com%2F20251210%2Fauto%2Fstorage%2Fgoog4_request&X-Goog-Date=20251210T122103Z&X-Goog-Expires=604800&X-Goog-SignedHeaders=host&X-Goog-Signature=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" ]	[ "Engagement" ]	wave1-1-LIVE NURSING INTERVIEW-Aj0nf3rKG_E_seg_003_1.mp4