Guangchen Lan

Senior Research Scientist

Palo Alto, CA

Hey! 👋 I am a Senior Research Scientist at Salesforce AI Research.

My research focuses on reinforcement learning (RL) for post-training foundation models, particularly large language models (LLMs). I work on RL with verifiable rewards (RLVR) as well as with non-verifiable rewards, such as rubric-based rewards and human feedback (RLHF). I currently focus on agentic RL and agent safety.

News

Sep 06, 2025	Hello, World! The new personal website has been launched.

Selected Publications

NeurIPS 2026

MaPPO: Maximum a Posteriori Preference Optimization with Prior Knowledge

Guangchen Lan, Sipeng Zhang, Tianle Wang, Yuwei Zhang, Daoan Zhang, Xinpeng Wei, and 4 more authors

(Featured by HuggingFace 🤗 Daily Papers) NeurIPS, 2026

Bib PDF

@article{lan2025mappo,
  title = {{MaPPO}: Maximum a Posteriori Preference Optimization with Prior Knowledge},
  author = {Lan, Guangchen and Zhang, Sipeng and Wang, Tianle and Zhang, Yuwei and Zhang, Daoan and Wei, Xinpeng and Pan, Xiaoman and Zhang, Hongming and Han, Dong-Jun and Brinton, Christopher G},
  journal = {(Featured by HuggingFace 🤗 Daily Papers) NeurIPS},
  year = {2026},
}

NeurIPS 2025

Contextual Integrity in LLMs via Reasoning and Reinforcement Learning

Guangchen Lan, Huseyin A Inan, Sahar Abdelnabi, Janardhan Kulkarni, Lukas Wutschitz, Reza Shokri, and 2 more authors

In NeurIPS, 2025

Bib PDF

@inproceedings{lan2025contextual,
  title = {Contextual Integrity in {LLMs} via Reasoning and Reinforcement Learning},
  author = {Lan, Guangchen and Inan, Huseyin A and Abdelnabi, Sahar and Kulkarni, Janardhan and Wutschitz, Lukas and Shokri, Reza and Brinton, Christopher G and Sim, Robert},
  booktitle = {NeurIPS},
  year = {2025},
}

ICLR 2025

Asynchronous Federated Reinforcement Learning with Policy Gradient Updates: Algorithm Design and Convergence Analysis

Guangchen Lan, Dong-Jun Han, Abolfazl Hashemi, Vaneet Aggarwal, and Christopher G Brinton

In ICLR, 2025

Bib PDF

@inproceedings{lan2025asynchronous,
  title = {Asynchronous Federated Reinforcement Learning with Policy Gradient Updates: Algorithm Design and Convergence Analysis},
  author = {Lan, Guangchen and Han, Dong-Jun and Hashemi, Abolfazl and Aggarwal, Vaneet and Brinton, Christopher G},
  booktitle = {ICLR},
  year = {2025},
}

NeurIPS 2023

Improved Communication Efficiency in Federated Natural Policy Gradient via ADMM-based Gradient Updates

Guangchen Lan, Han Wang, James Anderson, Christopher Brinton, and Vaneet Aggarwal

In NeurIPS, 2023

Bib PDF

@inproceedings{lan2023improved,
  title = {Improved Communication Efficiency in Federated Natural Policy Gradient via ADMM-based Gradient Updates},
  author = {Lan, Guangchen and Wang, Han and Anderson, James and Brinton, Christopher and Aggarwal, Vaneet},
  booktitle = {NeurIPS},
  year = {2023},
}

IEEE Trans

Communication-Efficient Federated Learning for Resource-Constrained Edge Devices

Guangchen Lan, Xiao-Yang Liu, Yijing Zhang, and Xiaodong Wang

IEEE Transactions on Machine Learning in Communications and Networking, 2023

Bib PDF

@article{lan2023communication,
  title = {Communication-Efficient Federated Learning for Resource-Constrained Edge Devices},
  author = {Lan, Guangchen and Liu, Xiao-Yang and Zhang, Yijing and Wang, Xiaodong},
  journal = {IEEE Transactions on Machine Learning in Communications and Networking},
  year = {2023},
}