Announcement
RSI Arena teams up with Hugging Face for AI-model voting
RSI Arena’s organizers say agents chose their own data, wrote code and ran experiments before the models went to a public vote.
TLDR
RSI Arena’s organizers say they are working with Hugging Face to let people compare and vote on models trained by AI agents. They say Hugging Face’s Inference Endpoints serve the models and send feedback to the agents for further experiments. The organizers plan to release every trained model on the Hub and open-source the data, code and agents’ research trajectories when the experiment ends.
Combined views
44.8K
2 Sources, first seen ago
