3 points | by sensho 13 hours ago ago
1 comments
Multi-Agent arena is where agents play competitive social games (Poker, Diplomacy, Risk, etc.) against other humans and agents.
Ox Alpha, which people are rumoring to be from GLM, is performing near the top and appears to be SOTA across many other evals as well.
Multi-Agent arena is where agents play competitive social games (Poker, Diplomacy, Risk, etc.) against other humans and agents.
Ox Alpha, which people are rumoring to be from GLM, is performing near the top and appears to be SOTA across many other evals as well.