paper-with-me

홈 › Papers

STEP: Success-Rate-Aware Trajectory-Efficient Policy Optimization

2025-11-17 · Yuhan Chen, Yuxuan Liu, Long Zhang, Pengzhi Gao, Jian Luan, Wei Liu arxiv

Multi-turn interaction remains challenging for online reinforcement learning. A common solution is trajectory-level optimization, which treats each trajectory as a single training sample. However, this approach can be inefficient and yield misleading learning signals: it applies uniform sampling across tasks regardless of difficulty, penalizes correct intermediate actions in failed trajectories, and incurs high sample-collection costs. To address these issues, we propose STEP (Success-rate-aware Trajectory-Efficient Policy optimization), a framework that dynamically allocates sampling based on per-task success rates and performs step-level optimization. STEP maintains a smoothed success-rate record to guide adaptive trajectory resampling, allocating more effort to harder tasks. It then computes success-rate-weighted advantages and decomposes trajectories into step-level samples. Finally, it applies a step-level GRPO augmentation to refine updates for low-success tasks. Experiments on OSWorld and AndroidWorld show that STEP substantially improves sample efficiency and training stability over trajectory-level GRPO, converging faster and generalizing better under the same sampling budget.

📄 PDF Abstract BibTeX arXiv:2511.13091

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Similar Papers 제목 키워드 기반

Graph-Enhanced Policy Optimization in LLM Agent Training

2025-10-30 · Jiazhen Yuan, Zhike Gong, Jinquan Hang, Zhengbiao Bai 외 arxiv

Multi-step LLM agents in interactive environments represent a crucial step toward long-horizon decision-making. To train such agents, group-based reinforcement learning is widely adopted, which reinforces trajectories wi…

Reinforcement LearningSemantic Similarity

Trajectory-Aware Eligibility Traces for Off-Policy Reinforcement Learning

2023-01-26 · Brett Daley, Martha White, Christopher Amato, Marlos C. Machado

Off-policy learning from multistep returns is crucial for sample-efficient reinforcement learning, but counteracting off-policy bias without exacerbating variance is challenging. Classically, off-policy bias is corrected…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)

STAPO: Selective Trajectory-Aware Policy Optimization for LLM Agent Training

2026-07-06 · Qiuyi Qi, Tian Liang, Mutian Bao, Jinjian Zhang 외 arxiv

Reinforcement Learning (RL) is the dominant paradigm for training Large Language Model (LLM) agents on long-horizon tasks. However, sparse and delayed rewards often lead to trajectory neglect, in which agents lose focus …

Reinforcement Learning

StepOPSD: Step-Aware Online Preference Self-Distillation for Agent Reinforcement Learning

2026-05-26 · Yanfei Zhang, Xu Lin, Chenglin Wu arxiv

Reinforcement learning for multi-turn agents suffers from a credit-assignment mismatch: rewards are sparse and trajectory-level, while success often hinges on a few local decisions. Existing online policy distillation (O…

Reinforcement Learning

LAGO Policy: Latency-Aware Asynchronous Diffusion Policies with Goal-Directed Collision-Free Planning for Smooth Manipulation

2026-06-16 · Guowei Shi, Xupeng Xie, Yiming Luo, Jian Guo 외 arxiv

Diffusion-based visuomotor policies deployed with asynchronous inference often exhibit inter-chunk discontinuities and lack explicit mechanisms for obstacle-aware execution, leading to jerky motions and collisions that h…

Trajectory Planning