From Reward-Hack Activations to Agentic Risk States: Context-Calibrated Mechanistic Monitoring in LLM Agents 事件

From Reward-Hack Activations to Agentic Risk States: Context-Calibrated Mechanistic Monitoring in LLM Agents arXiv:2606.06223v1 Announce Type: new Abstract: Language-model agents act through repeated cycles of observation, reasoning, and action selection, making safety monitoring depend on both internal model state and environment context. We study reward-hacking monitors in ReAct-style agents acting in Gameable ALFWorld and WebShop. Agents are instrumented with activation-based reward-hack sco