← Back to all news
arXiv AI By Qirui Liu, Wei Cao, Yu Mi, Yichen Sun, Xuekun Zhang, Lei Liu, Yue Shen, Zhixuan Chu

SCOPE: Selective Critical-Step On-Policy Self-Distillation for Agentic Reinforcement Learning

Read the original on arXiv AI →

SCOPE is a new framework for selective critical‑step on‑policy self‑distillation in agentic reinforcement learning. After a supervised warm‑up, the policy examines its own trajectories, generates hindsight skills for each step, and distills guidance only at selected error steps, using a bounded teacher‑student coefficient and a success‑relative pairwise length reward. Experiments on ALFWorld and WebShop show consistent performance gains across multiple model backbones, with Qwen2.5‑3B achieving 97.7 % success on ALFWorld and 85.2 % on WebShop, and Qwen3‑32B reaching 99.2 % on seen tasks and 94.0 % on unseen tasks.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

Related stories

arXiv AI
Aug 26

AHEAD: Adaptive Hindsight with Environment-Augmented Distillation for Agentic RL

AHEAD is a step‑aware framework that augments reinforcement learning for multi‑turn LLM agents by matching different supervision sources to different step types. The teacher receives environment feedback on all steps and LLM‑generated corrective hints only on error steps, providing finer‑grained guidance than uniform trajectory‑level rewards. Across ALFWorld, WebShop, and Search‑based QA, AHEAD improves task success by 13.3 points on ALFWorld and 11.0 on WebShop at 7B, reaches target success rates faster, and solves tasks within tighter interaction budgets compared to outcome‑only RL and prior self‑distillation baselines.

By Xiaolong Jin, Dingmin Wang, Vijay Lingam, Varun Kumar
llmsagentsreinforcement-learningefficiency
More like this →
arXiv AI
Jun 16

On-Policy Distillation with Curriculum Turn-level Guidance for Multi-turn Agents

arXiv:2606. 15912v1 Announce Type: cross Abstract: Multi-turn agents that plan, invoke tools, and interact with environments offer a promising paradigm for solving complex tasks, yet their capabilities typically rely on very large models whose inference cost is prohibitive in practice.

By Gengsheng Li, Mao Zheng, Mingyang Song, Ruiqi Liu, Tianyu Yang, Jie Sun, Qiyong Zhong, Haiyun Guo, Junfeng Fang, Dan Zhang, Jinqiao Wang
agentsefficiency
More like this →
arXiv AI
Aug 7

AgentOPSD: Recursive Self-Distillation for Agentic Reinforcement Learning

arXiv:2608. 05987v1 Announce Type: new Abstract: Reinforcement learning (RL) with verifiable rewards constructs trajectory-level advantage estimates, yet it often fails to credit the few pivotal decisions that determine outcomes in long-horizon, multi-turn agentic tasks.

By Zi-Han Wang, Zhengxi Lu, Zhiyuan Yao, Jinyang Wu, Jie Wu, Zhengzhou Cai, Yueqing Sun, Ziang Ye, Linji Hao, Qi Gu, Xunliang Cai, Yongliang Shen, Yujiu Yang
agentsreinforcement-learningefficiencybenchmarks
More like this →