Dataset for the leaderboard results of different LLMs

[Beliefuture]

Hi!

Are the results of different LLMs displayed on the leaderboard based on the dataset under the evaluator/dataset directory?
There appear to be only few records on the public dataset. If so, I would like to ask how such a small volume makes the results shown in the leaderboard convincing?

Could you provide more details about the dataset?

[iwanghc]

@Beliefuture We are in the process of expanding both the volume and the quality of our evaluation datasets, and will also include additional LLMs in our benchmark. Once the evaluations are complete, we will publish the updated dataset along with a refreshed leaderboard.