惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

Apple Machine Learning Research
Apple Machine Learning Research
J
Java Code Geeks
博
博客园 - 聂微东
Microsoft Azure Blog
Microsoft Azure Blog
量
量子位
T
Tailwind CSS Blog
Vercel News
Vercel News
I
InfoQ
Stack Overflow Blog
Stack Overflow Blog
U
Unit 42
Engineering at Meta
Engineering at Meta
L
LangChain Blog
大猫的无限游戏
大猫的无限游戏
D
Docker
博
博客园_首页
P
Proofpoint News Feed
月光博客
月光博客
T
The Blog of Author Tim Ferriss
MyScale Blog
MyScale Blog
酷 壳 – CoolShell
酷 壳 – CoolShell
Martin Fowler
Martin Fowler
腾
腾讯CDC
N
Netflix TechBlog - Medium
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More

cs.LG updates on arXiv.org

Memory-Guided Trust-Region Bayesian Optimization (MG-TuRBO) for High Dimensions EngageTriBoost: Predictive Modeling of User Engagement in Digital Mental Health Intervention Using Explainable Machine Learning Reservoir observer enhanced with residual calibration and attention mechanism Efficient RL Training for LLMs with Experience Replay Wireless Communication Enhanced Value Decomposition for Multi-Agent Reinforcement Learning Adversarial Sensor Errors for Safe and Robust Wind Turbine Fleet Control IKKA: Inversion Classification via Critical Anomalies for Robust Visual Servoing Adaptive Simulation Experiment for LLM Policy Optimization EvoLen: Evolution-Guided Tokenization for DNA Language Model Smartwatch-Based Sitting Time Estimation in Real-World Office Settings Structural Evaluation Metrics for SVG Generation via Leave-One-Out Analysis Loom: A Scalable Analytical Neural Computer Architecture Spectral Geometry of LoRA Adapters Encodes Training Objective and Predicts Harmful Compliance Finite-Sample Analysis of Nonlinear Independent Component Analysis:Sample Complexity and Identifiability Bounds How does Chain of Thought decompose complex tasks? Uncertainty-Aware Transformers: Conformal Prediction for Language Models Adaptive Candidate Point Thompson Sampling for High-Dimensional Bayesian Optimization Using Synthetic Data for Machine Learning-based Childhood Vaccination Prediction in Narok, Kenya Delve into the Applicability of Advanced Optimizers for Multi-Task Learning Bridging SFT and RL: Dynamic Policy Optimization for Robust Reasoning Multi-Agent Decision-Focused Learning via Value-Aware Sequential Communication Predictive Entropy Links Calibration and Paraphrase Sensitivity in Medical Vision-Language Models Efficient Hierarchical Implicit Flow Q-learning for Offline Goal-conditioned Reinforcement Learning Modality-Aware Zero-Shot Pruning and Sparse Attention for Efficient Multimodal Edge Inference The nextAI Solution to the NeurIPS 2023 LLM Efficiency Challenge Feature-Label Modal Alignment for Robust Partial Multi-Label Learning Integrated electro-optic attention nonlinearities for transformers Toward World Models for Epidemiology Tracing the Chain: Deep Learning for Stepping-Stone Intrusion Detection Batch Distillation Data for Developing Machine Learning Anomaly Detection Methods
Rollout Pass-Rate Control: Steering Binary-Reward RL Towa...
Tianshu Zhu, · 2026-05-07 · via cs.LG updates on arXiv.org

View PDF HTML (experimental)

Abstract:SWE-bench-style agentic reinforcement learning relies on expensive stateful trajectories, yet substantial compute is wasted on sampled rollout groups with skewed pass rates, where binary rewards provide a weak contrastive signal. We frame this inefficiency as a pass-rate control problem and show that a 50% pass rate is the most informative operating point: it maximizes reward entropy, the probability of surviving group filtering, RLOO advantage energy under GRPO, and success--failure contrastive structure. Guided by this principle, we propose Prefix Sampling (PS), which replays trajectory prefixes to steer skewed groups toward this regime: successful prefixes serve as head starts for mostly failing groups, while failing prefixes serve as handicaps for mostly passing groups. In stateful agent environments, prefix states are reconstructed through replay while replayed tokens are excluded from the loss, restricting optimization to continuations generated by the current policy. On SWE-bench-style agentic RL, PS delivers end-to-end wall-clock speedups of 2.01x on Qwen3-14B and 1.55x on Qwen3-32B while preserving or improving final verified performance. For 14B, the SWE-bench Verified peak rises from the baseline peak of 0.273 to 0.295 under PS. Additional mathematical reasoning experiments on AIME 2025 show the same pass-rate control pattern and decompose the gains into replay, bidirectional coverage, and adaptive control.
Comments: 25 pages, 8 figures, 11 tables
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2605.05112 [cs.LG]
  (or arXiv:2605.05112v1 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2605.05112

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Tianshu Zhu [view email]
[v1] Wed, 6 May 2026 16:44:38 UTC (469 KB)

此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。