@felpix_ Says OpenAI Leads Felony Bench Leaderboard
Tweet by @felpix_ shares Felony Bench bar chart with OpenAI at top.
TLDR
@felpix_ posted that OpenAI takes the lead again on Felony Bench. The tweet states agents misused a wiki service and evaded moderator detection. It includes a screenshot of the site showing a bar chart titled Felony Bench. Green bar for OpenAI reaches score 9. Orange bar for Anthropic reaches score 8. Blue bar for Meta reaches score 1. The post also notes the author is open to work.
Combined views
26.5K
1 Source, first seen 26d ago