Andy Su

RESEARCH SCIENTIST | NEW YORK, UNITED STATES

Andy Su is a research scientist at Meta FAIR focused on LLM reasoning. Andy received his PhD from Princeton University. Andy has a strong background in reinforcement learning and LLM.

Personal Website

Research Areas

Core Machine Learning

Natural Language Processing (NLP)

Reinforcement Learning

Andy's Work

Llama Flagship Models: Post-training, Reinforcement Learning, Reasoning, Scaling

LLM Reasoning with in-context search: Better Planning with Transformers via Search Dynamics Bootstrapping

LLM Latent Reasoning in Discrete Space: Mixing Latent and Text Tokens for Improved Language Model Reasoning

LLM Reasoning in Continuous Latent Space: Training Large Language Models to Reason in a Continuous Latent Space

LLM Reasoning: Dualformer-Controllable Fast and Slow Thinking by Learning with Randomized Reasoning Traces

Andy's Publications

October 13, 2025

REINFORCEMENT LEARNING

RESEARCH

SPG: Sandwiched Policy Gradient for Masked Diffusion Language Models

Chenyu Wang, Paria Rashidinejad, DiJia Su, Song Jiang, Sid Wang, Siyan Zhao, Cai Zhou, Shannon Zejiang Shen, Feiyu Chen, Tommi Jaakkola, Yuandong Tian, Bo Liu

October 13, 2025