Scholar

Ricky Tian Qi Chen

Google Scholar ID: 7MxQd6UAAAAJ

Meta FAIR

generative modelingdynamical systemsstochastic controlnormalizing flows

Citations & Impact

All-time

Citations

18,021

H-index

i10-index

Publications

Co-authors

list available

Contact

Publications

20 items

Browse publications on Google Scholar (top-right) ↗

Resume (English only)

Academic Achievements

2025, NeurIPS Oral: "Edit Flows: Flow Matching with Edit Operations".
2025, NeurIPS Oral: "Adjoint Schrödinger Bridge Sampler".
2025, ICML: "Adjoint Sampling: Highly Scalable Diffusion Samplers via Adjoint Matching".
2025, ICLR Oral: "Flow Matching with General Discrete Paths: A Kinetic-Optimal Perspective".
2025, ICLR Oral: "Generator Matching: Generative modeling with arbitrary Markov processes".
2025, ICLR Spotlight: "Adjoint Matching: Fine-tuning Flow and Diffusion Generative Models with Memoryless Stochastic Optimal Control".
2024, NeurIPS: "Discrete Flow Matching".
2024, ICML: "FlowMM: Generating Materials with Riemannian Flow Matching".
2024, ICLR: "Generalized Schrödinger Bridge Matching".
2024, ICLR Outstanding Paper Honorable Mention: "Flow Matching on General Geometries".
2023, ICLR Spotlight: "Flow Matching for Generative Modeling".

Background

Research Scientist at Meta.
Research focuses on building simplified abstractions of the world through the lens of dynamical systems and flows.
Recent work includes insertion-based sequence generation (Edit Flows) and scaling laws for multimodal generation (OneFlow).
Data-driven methods like Flow Matching have been widely adopted in video and audio foundation models (e.g., Movie Gen, SD3).
Reward-driven methods such as Adjoint Matching have been applied to large-scale diffusion fine-tuning and AI & Chemistry (e.g., AS/ASBS).

Co-authors

53 total