paper-with-me

홈 › Papers

PolicyBank: Evolving Policy Understanding for LLM Agents

2026-04-16 · Jihye Choi, Jinsung Yoon, Long T. Le, Somesh Jha, Tomas Pfister arxiv

LLM agents operating under organizational policies must comply with authorization constraints typically specified in natural language. In practice, such specifications inevitably contain ambiguities and logical or semantic gaps that cause the agent's behavior to systematically diverge from the true requirements. We ask: by letting an agent evolve its policy understanding through interaction and corrective feedback from pre-deployment testing, can it autonomously refine its interpretation to close specification gaps? We propose PolicyBank, a memory mechanism that maintains structured, tool-level policy insights and iteratively refines them -- unlike existing memory mechanisms that treat the policy as immutable ground truth, reinforcing "compliant but wrong" behaviors. We also contribute a systematic testbed by extending a popular tool-calling benchmark with controlled policy gaps that isolate alignment failures from execution failures. While existing memory mechanisms achieve near-zero success on policy-gap scenarios, PolicyBank closes up to 82% of the gap toward a human oracle.

📄 PDF Abstract BibTeX arXiv:2604.15505

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

APEX: Autonomous Policy Exploration for Self-Evolving LLM Agents

2026-05-20 · Yibo Li, Jiashuo Yang, Zhi Zheng, Zhiyuan Hu 외 arxiv

LLM agents have shown strong performance across a wide range of complex tasks, including interactive environments that require long-horizon decision making. But these agents cannot learn on the fly at test time. Self-evo…

Decision Making

Co-Evolution of Policy and Internal Reward for Language Agents

2026-04-03 · Xinyu Wang, Hanwei Wu, Jingwei Song, Shuyuan Zhang 외 arxiv

Large language model (LLM) agents learn by interacting with environments, but long-horizon training remains fundamentally bottlenecked by sparse and delayed rewards. Existing methods typically address this challenge thro…

NavMorph: A Self-Evolving World Model for Vision-and-Language Navigation in Continuous Environments

2025-06-30 · Xuan Yao, Junyu Gao, Changsheng Xu

Vision-and-Language Navigation in Continuous Environments (VLN-CE) requires agents to execute sequential navigation actions in complex environments guided by natural language instructions. Current approaches often strugg…

Decision MakingVision and Language Navigation

Self-Evolving Multi-Agent Network for Industrial IoT Predictive Maintenance

2026-02-17 · Rebin Saleh, Khanh Pham Dinh, Balázs Villányi, Truong-Son Hy arxiv

Industrial IoT predictive maintenance requires systems capable of real-time anomaly detection without sacrificing interpretability or demanding excessive computational resources. Traditional approaches rely on static, of…

Response GenerationAnomaly Detection

Self-evolving LLM agents with in-distribution Optimization

2026-06-05 · Yudi Zhang, Meng Fang, Zhenfang Chen, Mykola Pechenizkiy arxiv

Large Language Models (LLMs) have recently emerged as powerful controllers for interactive agents in complex environments, yet training them to perform reliable long-horizon decision making remains a fundamental challeng…

Reinforcement LearningDecision Making