Why Multi-Step Tool-Use Reinforcement Learning Collapses and How Supervisory Signals Fix It
Paper • 2606.26027 • Published • 15
Lei Wang
demolei
AI & ML interests
LLMs
Recent Activity
View all activity
Organizations
upvoted 5
papers 1 day ago
GUI vs. CLI: Execution Bottlenecks in Screen-Only and Skill-Mediated Computer-Use Agents
Paper • 2606.24551 • Published • 25
The Verification Horizon: No Silver Bullet for Coding Agent Rewards
Paper • 2606.26300 • Published • 37
OPID: On-Policy Skill Distillation for Agentic Reinforcement Learning
Paper • 2606.26790 • Published • 40
DanceOPD: On-Policy Generative Field Distillation
Paper • 2606.27377 • Published • 62
upvoted a paper 2 days ago
upvoted 5
papers 3 days ago
PlanBench-XL: Evaluating Long-Horizon Planning of LLM Tool-Use Agents in Large-Scale Tool Ecosystems
Paper • 2606.22388 • Published • 95
OpenThoughts-Agent: Data Recipes for Agentic Models
Paper • 2606.24855 • Published • 43
Qwen-AgentWorld: Language World Models for General Agents
Paper • 2606.24597 • Published • 132
MobileForge: Annotation-Free Adaptation for Mobile GUI Agents with Hierarchical Feedback-Guided Policy Optimization
Paper • 2606.19930 • Published • 42
MemGUI-Agent: An End-to-End Long-Horizon Mobile GUI Agent with Proactive Context Management
Paper • 2606.19926 • Published • 42
upvoted a paper 12 days ago
upvoted 4
papers 15 days ago
EvoBrowseComp: Benchmarking Search Agents on Evolving Knowledge
Paper • 2606.13120 • Published • 4
WeaveBench: A Long-Horizon, Real-World Benchmark for Computer-Use Agents with Hybrid Interfaces
Paper • 2606.09426 • Published • 104
FORT-Searcher: Synthesizing Shortcut-Resistant Search Tasks for Training Deep Search Agents
Paper • 2606.12087 • Published • 77
EvoArena: Tracking Memory Evolution for Robust LLM Agents in Dynamic Environments
Paper • 2606.13681 • Published • 142
upvoted a paper 18 days ago
upvoted a paper 22 days ago
upvoted 2
papers 24 days ago