Image de couverture de livre

Does RL Expand the Capability Boundary of LLM Agents? A PASS@(k,T) Analysis

Fuente: arXiv

Enregistré dans:

Détails bibliographiques
Auteurs principaux:	Zhai, Zhiyuan, Yan, Wenjing, Shao, Xiaodan, Wang, Xin
Format:	Preprint
Publié:	2026
Sujets:	Machine Learning
Accès en ligne:	Acceder al recurso
Tags:	Ajouter un tag Pas de tags, Soyez le premier à ajouter un tag!

Documents similaires

Selective Rollout: Mid-Trajectory Termination for Multi-Sample Agent RL
par: Zhai, Zhiyuan, et autres
Publié: (2026)

How Much Thinking is Enough? Quantifying and Understanding Redundancy in LLM Reasoning
par: Zhai, Zhiyuan, et autres
Publié: (2026)

Revisable by Design: A Theory of Streaming LLM Agent Execution
par: Zhai, Zhiyuan, et autres
Publié: (2026)

Expanding LLM Agent Boundaries with Strategy-Guided Exploration
par: Szot, Andrew, et autres
Publié: (2026)

When Does Multi-Agent RL Improve LLM Workflows? Workflow, Scale, and Policy-Sharing Tradeoffs
par: Zeng, Yifan, et autres
Publié: (2026)

Normalizing Flows are Capable Models for RL
par: Ghugare, Raj, et autres
Publié: (2025)

SPA-RL: Reinforcing LLM Agents via Stepwise Progress Attribution
par: Wang, Hanlin, et autres
Publié: (2025)

RL-PLUS: Countering Capability Boundary Collapse of LLMs in Reinforcement Learning with Hybrid-policy Optimization
par: Dong, Yihong, et autres
Publié: (2025)

CoBA-RL: Capability-Oriented Budget Allocation for Reinforcement Learning in LLMs
par: Yao, Zhiyuan, et autres
Publié: (2026)

Towards Mitigation of Hallucination for LLM-empowered Agents: Progressive Generalization Bound Exploration and Watchdog Monitor
par: Liu, Siyuan, et autres
Publié: (2025)

Quantifying the Capability Boundary of DeepSeek Models: An Application-Driven Performance Analysis
par: Zhao, Kaikai, et autres
Publié: (2025)

SWEET-RL: Training Multi-Turn LLM Agents on Collaborative Reasoning Tasks
par: Zhou, Yifei, et autres
Publié: (2025)

JigsawRL: Assembling RL Pipelines for Efficient LLM Post-Training
par: Hu, Zhengding, et autres
Publié: (2026)

Memex(RL): Scaling Long-Horizon LLM Agents via Indexed Experience Memory
par: Wang, Zhenting, et autres
Publié: (2026)

Critique to Verify: Accurate and Honest Test-Time Scaling with RL-Trained Verifiers
par: Yang, Zhicheng, et autres
Publié: (2025)

PASS: Private Attributes Protection with Stochastic Data Substitution
par: Chen, Yizhuo, et autres
Publié: (2025)

LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage
par: Nie, Yuzhou, et autres
Publié: (2024)

LLM Augmented LLMs: Expanding Capabilities through Composition
par: Bansal, Rachit, et autres
Publié: (2024)

Expanding the Capabilities of Reinforcement Learning via Text Feedback
par: Song, Yuda, et autres
Publié: (2026)

Adaptive Test-Time Compute Allocation for Reasoning LLMs via Constrained Policy Optimization
par: Zhai, Zhiyuan, et autres
Publié: (2026)

CurveRL: Principled Distribution-Aware Context Reweighting for LLM Reasoning
par: Sun, Ke, et autres
Publié: (2026)

Evaluating PhaseNet on Teleseismic Data with MsPASS
par: Ma, Jinxin, et autres
Publié: (2026)

Problem-Parameter-Free Decentralized Bilevel Optimization
par: Zhai, Zhiwei, et autres
Publié: (2025)

ArenaRL: Scaling RL for Open-Ended Agents via Tournament-based Relative Ranking
par: Zhang, Qiang, et autres
Publié: (2026)

RL Grokking Recipe: How Does RL Unlock and Transfer New Algorithms in LLMs?
par: Sun, Yiyou, et autres
Publié: (2025)

Paying Less Generalization Tax: A Cross-Domain Generalization Study of RL Training for LLM Agents
par: Liu, Zhihan, et autres
Publié: (2026)

MATH-Beyond: A Benchmark for RL to Expand Beyond the Base Model
par: Mayilvahanan, Prasanna, et autres
Publié: (2025)

Assessing the Zero-Shot Capabilities of LLMs for Action Evaluation in RL
par: Pignatelli, Eduardo, et autres
Publié: (2024)

MemAgent: Reshaping Long-Context LLM with Multi-Conv RL-based Memory Agent
par: Yu, Hongli, et autres
Publié: (2025)

CodeARC: Benchmarking Reasoning Capabilities of LLM Agents for Inductive Program Synthesis
par: Wei, Anjiang, et autres
Publié: (2025)

RL in the Wild: Characterizing RLVR Training in LLM Deployment
par: Zhou, Jiecheng, et autres
Publié: (2025)

Enhancing RL Safety with Counterfactual LLM Reasoning
par: Gross, Dennis, et autres
Publié: (2024)

TSR: Trajectory-Search Rollouts for Multi-Turn RL of LLM Agents
par: Djuhera, Aladin, et autres
Publié: (2026)

Multi-Agent Path Finding via Offline RL and LLM Collaboration
par: Atasever, Merve, et autres
Publié: (2025)

How Does RL Post-training Induce Skill Composition? A Case Study on Countdown
par: Park, Simon, et autres
Publié: (2025)

How does Your RL Agent Explore? An Optimal Transport Analysis of Occupancy Measure Trajectories
par: Nkhumise, Reabetswe M., et autres
Publié: (2024)

Test-time RL alignment exposes task familiarity artifacts in LLM benchmarks
par: Wang, Kun, et autres
Publié: (2026)

Sundial: A Family of Highly Capable Time Series Foundation Models
par: Liu, Yong, et autres
Publié: (2025)

Compute-Optimal Scaling for Value-Based Deep RL
par: Fu, Preston, et autres
Publié: (2025)

PASS: Probabilistic Agentic Supernet Sampling for Interpretable and Adaptive Chest X-Ray Reasoning
par: Feng, Yushi, et autres
Publié: (2025)