Agents
Language models that use tools, learn from environments, and improve through experience.
Ning Lu · AI Researcher
I teach language agents
to think, act, and imagine.
My research explores agentic reinforcement learning and world models— building AI systems that learn through interaction, reason over long horizons, and remain aligned along the way.
A research philosophy
Intelligence is not only about producing an answer. It is about understanding what happens next.
Worlds
Models that anticipate consequences and turn imagination into better decisions.
Alignment
Methods that keep capable models safe, useful, and faithful as they adapt.
Selected research
Ideas, made concrete.
Recent work across agent learning, reasoning, world modeling, and language-model safety.
Policy and World Modeling Co-Training for Language Agents
The first policy and world-modeling co-training RL framework for LLM agents.
Train at the Moving Edge: Efficient RL for Large Reasoning Models via Rollout Selection
The first online policy-verified data selection framework for efficient RL training.
Is PRM Necessary? Problem-Solving RL Implicitly Induces PRM Capability in LLMs
Unifying problem solving and solution-process judgment.
About Ning
At the boundary of
reasoning and action.
I received my Ph.D. in CSE from the Hong Kong University of Science and Technology (HKUST), where I was supervised by Prof. Cunsheng Ding, Prof. Qi Wang, and Prof. Ke Tang and focused on trustworthy language models. Previously I was an intern researcher at Bytedance-Seed, where I focus on the Agentic Reinforcement Learning. Now I am focusing on Agentic Reinforcement Learning and World Modeling for LLM.
Research Interest:
- LLM RL for Agent and Reasoning: PaW, HIVE, AHDAgent, Is PRM Necessary?
- LLM Alignment & RedTeaming: SafeDelta, SICO
DegreePh.D. in CSE
Alma materHKUST
Current lensAgents × Worlds