Tsinghua's AAArena: Agents Learn Games by Post-Match Review, Opus5.5 Wins Half the Golds

Tsinghua introduced AAArena, a benchmark of 12 adversarial games with 1,920 historical human bots, plus the AHL paradigm where weight-frozen agents improve by reading rules and reviewing matches; Opus5.5 took half the gold medals.

2026-10-11 ~ 2026-10-12 · 2 related posts

• Tsinghua's AAArena finds AI agents hit rank 1 with replays but stall on complex games — rohanpaul_ai · 2026-10-11
• Tsinghua's AAArena: Claude Opus 5.5 + Claude Code Wins Half of 12 Adversarial Game Golds — THU-CST-SAST · 2026-10-12