MINAKANUSHI-6.8B

Status:              Research checkpoint
Training cycle:      v0.3.1
Capability verdict:  pending compare_v031.py
Accepted:            NO
Not a language model
Canonical runtime:   .mina
Safetensors:         weight mirror only
Action output:       ActionIntent
PWM:                 false

v0.3.1 is not accepted yet. Latest weights are a research checkpoint after 1000 H200 steps from step128. Train-eval hint is B / C-signal. Official A/B/C comes only from scripts/compare_v031.py on the capability ledger (heldout 100, revision, false revision, direction, memory on/off ADE, WAIT vs MOVE_TO, action learning). Do not read loss ↓ as a pass.

NULLXES MINAKANUSHI — adaptive situational intelligence for autonomous physical systems.

architecture: MINAKANUSHI
short name:   MINA
organization: NULLXES
author:       MagistrTheOne
parameters:   6 799 130 646
latent_dim:   4096
state_dim:    4096
memory_dim:   4096
core_depth:   32
world_slots:  512
memory_slots: 1024
native unit:  MinaUnit

Not a chatbot. Not a VLA. Not a wrapper around Qwen / Llama / Mistral / Gemma / DeepSeek / GPT / Claude.

Code: MagistrTheOne/NULLXES-MINAKANUSHI

---

Model parameters (frozen profile minakanushi_6_8b)

Formula inventory: 6 799 130 646 parameters. Architecture freeze 7aba976. Do not change these to “make it train”.

Parameter	Value
profile_name	minakanushi_6_8b
short_name	MINA
latent_dim / state_dim / memory_dim	4096
core_depth (DWC)	32
world_slots	512
memory_slots	1024
uncertainty_channels	8
dropout	0.0
future_branches	3
cognition.budget	4
cognition.convergence_threshold	0.02
prediction_horizons	immediate 1 · short 4 · medium 8
dt	0.1 s
max_sources / max_observations	64 / 64
NPF num_frequencies	32
persistence steps / retire	8 / 0.95
weights infer bf16	~13.6 GB
train	FSDP2 ZeRO-3, bf16 compute, AdamW

Identity (checkpoint metadata, not a prompt):

architecture: MINAKANUSHI
organization: NULLXES
system_class: adaptive_situational_intelligence
architecture_generation: 1
native_runtime: nullxes
architecture_version: "0.1"

---

Checkpoints (repo root)

Canonical runtime is *.mina. Every published model after v0.3.1 also ships a safetensors weight mirror. Safetensors is not the runtime.

File	What
minakanushi_stage0_step1128.mina	v0.3.1 research · 1× H200 · resume step128 · dataset/mina_6_8b_v03 · steps 129–1128 · not accepted
minakanushi_stage0_step128.mina	Status Core v0.2 · 1× B300 · resume IdentityBound · JSON curriculum · steps 65–128 · git ede6bda
minakanushi_stage0_step64.mina	Status Core v0.1 · 1× H200 · seed 11 · git d70bfc0 · 64 steps
model-00001-of-00003.safetensors … 00003	Weight mirror of step1128 only · bf16 · Hub badge
metrics_v031.jsonl · experiment_v031.jsonl	v0.3.1 train/eval log. Not a capability ledger.

Format: native nullxes-minakanushi zip (ZIP_STORED). Same architecture. v0.3.1 is a continuation, not a clone.

Do not convert step64. See docs/HF_SAFETENSORS_MIRROR.md.

Also at root: metrics_v02.jsonl, train_v02.log, reference_inference_v031.pt.

observation → MinaUnit → NullxesPositionField → WorldState
  → DynamicWorldCore → memory + uncertainty → situation
  → futures → strategies → ConstraintKernel → ActionIntent

MINA emits ActionIntent. Never raw PWM.

---

Checklist

Freeze: do not add layers / MoE / language head / identity_loss. Do not train 6.8B on CPU, RTX PRO 6000, or 1× H100 80GB.

Closed

• Gate 09 Runtime — observe → intent → restore (cpu_dev)
• Stage A GPU — 6.2M CUDA/bf16/AMP/.mina (stack, not intelligence)
• Gate 03B hidden direction — n=1000 · detected 0.986 · false_revision 0.0
• Status Core v0.1 step 64 — 1× H200 FSDP2 bf16
• Identity Initialization — passport stamp, no identity_loss
• JSON curriculum 1000 — physics/agency/causality/embodiment × 250 · pwm=false
• Resume v0.2 — 1× B300 · same model · dataset/mina_6_8b in the loss · steps 65–128

Open

• compare_v031.py — heldout 100 · revision · false revision · direction · memory on/off ADE · WAIT vs MOVE_TO · action. This is the A/B/C gate. Not v0.4.
• Long 6.8B — same freeze, only after compare verdict
• Gate 9+ perception — pixels → MinaUnit
• MINA V2 MM — organs → MinaUnit (not VLA/Cosmos)

Must-hold

constraint_violation_count == 0
persistence / reacquisition stay high
false_revision_rate stays low
hard constraints beat higher-value illegal strategies
ActionIntent ≠ PWM

---

Measured signals

v0.1 · step 64 · H200

Signal	Value
loss	78.36
future ADE / FDE	3.42 / 0.81
world position error	1.07
uncertainty calibration	0.38
persistence / reacquisition	1.0 / 1.0
constraint_violation_count	0
closed_loop_success_rate	1.0
false_revision_rate	0.0

v0.3.1 · step 1128 · H200 (research, not accepted)

Train-eval series only. Not compare_v031.py. Hint: B / late C-signal.

Signal	Value
heldout ADE (single-episode protocol)	1.61 @150 → 0.15 @750 → 0.44 @1100
false_revision	0 until eval 1100 (1.0 after step-1080 spike)
counterfactual distance	≈ 0.0008 entire run (WAIT ≈ MOVE_TO)
action term	≈ 0.499 frozen
revision_detection	often > 0 (was 0 at unexpected_stop on step128)
revision_direction	≈ 0 most batches
memΔ	latent L2, not ADE(on) < ADE(off)
capability verdict	pending compare_v031.py

Do not treat this table as v0.3.1 PASS.

v0.2 · step 128 · B300 (JSON resume)

Signal	Value
loss	41.10
step time (steady)	fwd ~1.42 s · bwd ~0.58 s
persistence / reacquisition	1.0 / 1.0
constraint_violation_count	0
closed_loop_success_rate	1.0
false_revision_rate	0.0
future ADE / FDE	2.05 / 0.68
world position error	0.55
uncertainty calibration	0.19
revision_accuracy	0.0
branch_coverage	0.0

Loss on mixed JSON phases is not the architecture gate. Revision accuracy and branch coverage are still not a pass. Do not add layers.

---

Load

from minakanushi.architecture.config import load_architecture
from minakanushi.architecture.model import MinakanushiSystem
from minakanushi.training.checkpoint import load_mina

arch = load_architecture("configs/architecture/minakanushi_6_8b.yaml")
system = MinakanushiSystem(arch)  # GPU-class machine
manifest = load_mina("minakanushi_stage0_step128.mina", system)

Resume (B300 / 2× H200, not a laptop):

torchrun --nproc_per_node=1 scripts/train.py \
  --config configs/training/mina_6_8b_v02.yaml \
  --out experiments/mina_6_8b_v02 \
  --resume minakanushi_stage0_step128.mina

---

Limits

• Status Researched. 64 + 64 synthetic steps. Not locomotion. Not vision-foundation trained.
• Do not treat this checkpoint as a chat model.

License

NULLXES MINAKANUSHI Research License. Research use. Redistribution as another model family, or as an LLM wrapper with this name, is not granted.

NULLXES                 organization / architecture owner
MagistrTheOne           author, HF namespace, runtime repo
MINAKANUSHI / MINA      architecture family / short name

I WILL SURVIVE. NULLXES.