# LLM Agents Overreact, Fail to Learn in Group Games

- slug: llm-agents-overreact-fail-to-learn-in-group-games
- date: 2026-04-06
- category: Agentics

Human groups learned from feedback and stabilized across games. LLM groups did not — reacting at nearly double the human rate to the same error signal, and never once holding a guess constant.

---