Questions about this leaderboard:
#2
by htkong - opened
I noticed there are two leaderboards related to Open Telco on Hugging Face:
https://huggingface.co/datasets/GSMA/leaderboard and https://huggingface.co/spaces/otellm/leaderboard
(1) Do these two leaderboards use the same testing methodology? Are they both tested with the same configuration of Inspect AI?
(2) For the same model, such as OpenAI/gpt-5, the metrics differ quite a lot between the two leaderboards. What could be the reason for this?
(3) Is it possible for us to upload the test results of other models that we have tested ourselves?