paper-with-me

홈 › Papers

Agentic Policy Optimization via Instruction-Policy Co-Evolution

2025-12-01 · Han Zhou, Xingchen Wan, Ivan Vulić, Anna Korhonen arxiv

Reinforcement Learning with Verifiable Rewards (RLVR) has advanced the reasoning capability of large language models (LLMs), enabling autonomous agents that can conduct effective multi-turn and tool-integrated reasoning. While instructions serve as the primary protocol for defining agents, RLVR typically relies on static and manually designed instructions. However, those instructions may be suboptimal for the base model, and the optimal instruction may change as the agent's policy improves and explores the interaction with the environment. To bridge the gap, we introduce INSPO, a novel Instruction-Policy co-evolution framework that integrates instruction optimization as a dynamic component of the reinforcement learning (RL) loop. INSPO maintains a dynamic population of instruction candidates that are sampled with questions, where reward signals in RL loops are automatically attributed to each instruction, and low performers are periodically pruned. New instructions are generated and verified through an on-policy reflection mechanism, where an LLM-based optimizer analyzes past experience from a replay buffer and evolves more effective strategies given the current policy. We conduct extensive experiments on multi-turn retrieval and reasoning tasks, demonstrating that INSPO substantially outperforms strong baselines relying on static instructions. INSPO discovers innovative instructions that guide the agent toward more strategic reasoning paths, achieving substantial performance gains with only a marginal increase in computational overhead.

📄 PDF Abstract BibTeX arXiv:2512.01945

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

On-Policy Self-Evolution via Failure Trajectories for Agentic Safety Alignment

2026-05-12 · Bo Yin, Qi Li, Xinchao Wang arxiv

Tool-using LLM agents fail through trajectories rather than only final responses, as they may execute unsafe tool calls, follow injected instructions, comply with harmful requests, or over-refuse benign tasks despite pro…

EVOM: Agentic Meta-Evolution of Actor-Critic Architectures for Reinforcement Learning

2026-06-24 · Boyun Zhang, Chao Wang, Kai Wu arxiv

In actor-critic reinforcement learning, network architectures are typically manually designed. Automating this design is challenging because each candidate must be trained before evaluation, and the design space is open-…

Reinforcement Learning

ReSkill: Reconciling Skill Creation with Policy Optimization in Agentic RL

2026-06-01 · Zelin He, Haotian Lin, Boran Han, Wei Zhu 외 arxiv

Agentic reinforcement learning (RL) enables LLM agents to improve continuously from environment rewards, yet the resulting policies do not systematically accumulate reusable strategies that generalize across tasks. Modul…

Reinforcement Learning

RAPO: Expanding Exploration for LLM Agents via Retrieval-Augmented Policy Optimization

2026-03-03 · Siwei Zhang, Yun Xiong, Xi Chen, Zi'an Jia 외 arxiv

Agentic Reinforcement Learning (Agentic RL) has shown remarkable potential in large language model-based (LLM) agents. These works can empower LLM agents to tackle complex tasks via multi-step, tool-integrated reasoning.…

Reinforcement Learning

AT$^2$PO: Agentic Turn-based Policy Optimization via Tree Search

2026-01-08 · Zefang Zong, Dingwei Chen, Yang Li, Qi Yi 외 arxiv

LLM agents have emerged as powerful systems for tackling multi-turn tasks by interleaving internal reasoning and external tool interactions. Agentic Reinforcement Learning has recently drawn significant research attentio…

Reinforcement Learning