Huggingface - Community Evals: Because we're done trusting black-box leaderboards over the community

This blog post explains how Hugging Face is democratizing artificial intelligence evaluation by decentralizing and making evaluation results more transparent. Benchmark datasets on Hugging Face can now host leaderboards, and models can store their own evaluation scores. The community can submit results via PR, and verified badges prove that the results can be reproduced. This initiative aims to expose existing scores and make the Hub an active place to build and share reproducible benchmarks.