About

I am a Research Scientist at Luma AI, where I work on post-training and agentic systems for multimodal and physical AI. My current work includes an agent-as-policy framework for robotic manipulation and a multimodal design agent for infographic generation.

Previously, I worked on Foundation AI at LinkedIn, focusing on cross-episode meta-RL, tool-use search agents, preference optimization, on-policy distillation, and efficient production LLM systems. I am especially interested in LLM agents, post-training, and efficient model training and inference.

News

  • Joined Luma AI to work on foundation models and physical AI.
  • Released Cross-Episode Meta-RL for LLM in-context online learning.
  • Contributed to open-sourcing FMChisel, a research library for foundation-model compression and distillation.
  • AlphaPO appeared at ICML 2025.

Selected Work

A multimodal AI system transforming visual and text inputs into HTML code and a polished infographic

Multimodal Design Agent

Luma AI · Generative models · Reward modeling

Post-trained a VLM with RL to generate polished HTML infographics and developed a reward model aligned with human aesthetic preferences.

Diagram comparing isolated RL training with multi-episode meta-RL

Cross-Episode Online Learning

Meta-RL

Improved how LLMs learn from context and adapt across episodes using cross-episode reinforcement learning.

Paper

A search agent traversing connected information sources, calling tools, synthesizing an answer, and applying rubric-based evaluation

Tool-Use Search Agent

Synthetic data · Agent evaluation · Agentic RL

Graph-based query and answer generation, rubric-based evaluation, and agentic RL training in verl.

FMChisel GitHub repository preview

FMChisel

LinkedIn · LLM efficiency · Open source

Contributor to an open-source research library for efficient foundation-model algorithms, including compression, distillation, pruning, quantization, and efficient training.

GitHub

Selected Publications

  1. 2026
    Pretraining Knows What, Simulation Teaches How: Compositional Generalization for Agentic Robotics

    Benjamin Pikus, Daniel Saunders, Mircea Mironenco, Salman Mohammadi, Mathew Mathew, Sirou Zhu, Divya Makkar, Xuhua Huang, Qinbo Li, Supreeth Narasimhaswamy, Siyuan Cen, Dejia Xu, Shyamal Buch, Thomas Neff

    CoRL 2026 Late Breaking Results
  2. 2026
    Scaling In-Context Online Learning Capability of LLMs via Cross-Episode Meta-RL

    Xiaofeng Lin, Sirou Zhu, Yilei Chen, Mingyu Chen, Hejian Sang, Ioannis Paschalidis, Zhipeng Wang, Aldo Pacchiano, Xuezhou Zhang

    Preprint
  3. 2025
    Scaling Down, Serving Fast: Compressing and Deploying Efficient LLMs for Recommendation Systems

    Kayhan Behdin, Ata Fatahibaarzi, Qingquan Song, Yun Dai, Aman Gupta, Zhipeng Wang, Hejian Sang, Shao Tang, Gregory Dexter, Sirou Zhu, Siyu Zhu, Tejas Dharamsi, Vignesh Kothapalli, Zhoutong Fu, Yihan Cao, Pin-Lun Hsu, Fedor Borisyuk, Natesh S. Pillai, Luke Simon, Rahul Mazumder

    EMNLP Industry
  4. 2025
    AlphaPO: Reward Shape Matters for LLM Alignment

    Aman Gupta, Shao Tang, Qingquan Song, Sirou Zhu, Jiwoo Hong, Ankan Saha, Viral Gupta, Noah Lee, Eunki Kim, Siyu Zhu, Parag Agrawal, Natesh Pillai, S. Sathiya Keerthi

    ICML
  5. 2024
    LiGNN: Graph Neural Networks at LinkedIn

    Fedor Borisyuk, Shihai He, Yunbo Ouyang, Morteza Ramezani, Peng Du, Xiaochen Hou, Chengming Jiang, Nitin Pasumarthy, Priya Bannur, Birjodh Tiwana, Ping Liu, Siddharth Dangi, Daqi Sun, Zhoutao Pei, Xiao Shi, Sirou Zhu, Qianqi Shen, Kuang-Hsuan Lee, David Stein, Baolei Li, Haichao Wei, Amol Ghoting, Souvik Ghosh

    KDD Best Paper
  6. 2024
    LiRank: Industrial Large-Scale Ranking Models at LinkedIn

    Fedor Borisyuk, Mingzhou Zhou, Qingquan Song, Siyu Zhu, Birjodh Tiwana, Ganesh Parameswaran, Siddharth Dangi, Lars Hertel, Qiang Xiao, Xiaochen Hou, Yunbo Ouyang, Aman Gupta, Sheallika Singh, Dan Liu, Hailing Cheng, Lei Le, Jonathan Hung, Sathiya Keerthi, Ruoyan Wang, Fengyu Zhang, Mohit Kothari, Chen Zhu, Daqi Sun, Yun Dai, Xun Luan, Sirou Zhu, Zhiwei Wang, Neil Daftary, Qianqi Shen, Chengming Jiang, Haichao Wei, Maneesh Varshney, Amol Ghoting, Souvik Ghosh

    KDD

Experience

Luma AI
Research Scientist, Foundation Models

LinkedIn
Senior / Staff Research Engineer, Foundation AI

LinkedIn
Machine Learning Engineer, Ads Ranking

Education

Carnegie Mellon University
M.S. in Computational Data Science, 2021

Zhejiang University
B.E. in Software Engineering, 2019