LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents
Paper • 2608.17393 • Published • 22
Lei Wang
demolei
AI & ML interests
LLMs
Recent Activity
upvoted a paper about 7 hours ago
LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents upvoted a paper 7 days ago
An AI4AI Framework for Visual Token Pruning upvoted a paper 7 days ago
AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic DesignOrganizations
upvoted a paper about 7 hours ago
upvoted 4
papers 7 days ago
An AI4AI Framework for Visual Token Pruning
Paper • 2608.07193 • Published • 15
AutoDesign: Meta-Harness Optimization for Long-Horizon Agentic Design
Paper • 2608.13560 • Published • 53
DarwinX: Evolving Agent Harnesses Through Natural Selection
Paper • 2608.07545 • Published • 108
LLMRouter: Unified Infrastructure for Developing, Evaluating, and Deploying LLM Routers
Paper • 2608.06867 • Published • 107
upvoted 3
papers 8 days ago
OpenART: Scaling Agent Red Teaming via Open-Ended Environment Evolution
Paper • 2608.00677 • Published • 260
Co-Evolution in Agentic Systems: Toward Self-Directed Evolution Beyond Human Design
Paper • 2608.10299 • Published • 132
AI4AI at Test-Time: Strong-to-Weak Capability Transfer via Harnesses
Paper • 2608.12307 • Published • 111
upvoted 5
papers 9 days ago
DeepSearch-World: Self-Distillation for Deep Search Agents in a Verifiable Environment
Paper • 2607.07820 • Published • 94
AREX: Towards a Recursively Self-Improving Agent for Deep Research
Paper • 2607.21461 • Published • 154
BDH-CQ: In-Context Learning with Recurrent Latent Reasoning
Paper • 2608.09888 • Published • 692
Macaron-V1: Towards Open Continual Learning with Self-Improvement and Mixture-of-LoRA
Paper • 2608.09819 • Published • 336
Mendel Gödel Machine: Recursive Self-Improving Coding Agents via Comparative Evolution
Paper • 2608.07645 • Published • 26
upvoted 7
papers 14 days ago
AgentOPSD: Recursive Self-Distillation for Agentic Reinforcement Learning
Paper • 2608.05987 • Published • 98
OSReward: Instituting Standardized Evaluation for Cross-Platform Computer-Use Reward Models
Paper • 2607.28609 • Published • 72
Dr. Zero: Self-Evolving Search Agents without Training Data
Paper • 2601.07055 • Published • 23
Teaching Models to Teach Themselves: Reasoning at the Edge of Learnability
Paper • 2601.18778 • Published • 44
Recursive Harness Self-Improvement
Paper • 2607.15524 • Published • 26
SkillHone: A Harness for Continual Agent Skill Evolution Through Persistent Decision History
Paper • 2606.08671 • Published • 47
EvoClawBench: Can Agents Learn Reusable Skills from Their Own Runs?
Paper • 2607.09711 • Published • 1