Research &
publications.

LM2LM

Language Models Learn to Improve Language Models

Minhua Lin, Zijun Wang, Zewen Liu, et al.

Learn a proposal policy from RWM rewards calibrated by real experiments.

Research World Models

Language Models as AI Research World Models

Zijun Wang, Zewen Liu, Minhua Lin, et al.

Predict intervention outcomes using reusable experimental experience.

Evo-Harness

Context-to-Harness Skill Compilation for Self-Evolving Agents

Tianxin Wei, Zhan Shi, et al.

Compile noisy execution context into reusable skills.

Harness Disentanglement

Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agents

Minhua Lin, Juncheng Wu, et al.

Separate the ability to produce harness updates from the ability to benefit from them.

MemMA

Coordinating the Memory Cycle through Multi-Agent Reasoning and In-Situ Self-Evolution

Minhua Lin, Zhiwei Zhang, Hanqing Lu, et al.

Coordinate memory construction and retrieval with verification and repair.

Adaptive Auto-Harness

Sustained Self-Improvement for Agentic System Deployment on Open-Ended Task Streams

Zewen Liu, Zhan Shi, et al.

Adapt harnesses and route tasks as deployment conditions change.

A-Evolve-Training

Autonomous Post-Training of a 30B Model

Zhan Shi, Bing He, Yisi Sang, Benoit Dumoulin, Hanqing Lu

An autonomous system manages a multi-week model-development campaign.

Agentic Evolution

Agentic Evolution is the Path to Evolving LLMs

A research position on improving the agent system around a model.