一作ACL 2026Findings· Findings of ACL
LCO : LLM -based Constraint Optimization for Safer Agentic LLM s in Real-world Tasks
Jiayong Wan · Jiawei Chen · Zhaoxia Yin · Liu Shuyuan · Hang Su
摘要
Large Language Models (LLMs) are increasingly acting as autonomous agents, but their continuous interaction with the environment can lead to in-context reward hacking (ICRH), a phenomenon in which LLMs iteratively optimize their behavior to maximize proxy objectives, inadvertently producing harmf...