Ruibo Chen

About

I am a fourth-year Ph.D. student in Computer Science at the University of Maryland, College Park, advised by Heng Huang, Brendan Iribe Endowed Professor of Computer Science. Previously I received my B.S. in Intelligence Science and Technology from the School of Electronics Engineering and Computer Science at Peking University, where I was fortunate to work with Prof. Xu Sun.

My research is about making generative models more capable and, just as importantly, more trustworthy. On the capability side, I work on data selection, preference optimization, and test-time scaling for vision-language understanding and image generation. On the trustworthiness side, I design distortion-free watermarks that carry provenance signals at no cost to generation quality, across text, speech, audio, and protein sequences, and I study attacks such as watermark removal, since robustness against an adversary is what makes these guarantees usable in practice.

I am always happy to connect. Feel free to email me at rbchen [at] umd [dot] edu if you are interested in my research or possible collaborations.

I am actively looking for research internship opportunities for Spring and Summer 2027!

News

  • Oct 2026 Glad to see that OpenAI’s textGrain watermark, an improved variant of our proposed MCMark (ACL 2025), is being deployed in ChatGPT and Codex!
  • Sep 2026 MC2Mark accepted to NeurIPS 2026.
  • Aug 2026 More Haste, Less Speed accepted to EMNLP 2026 (Main Conference).
  • Jun 2026 POET accepted to ECCV 2026.
  • May 2026 Research Scientist Intern at TikTok, developing the core watermarking algorithms for ByteDance LLM products.

Selected Publications

* denotes equal contribution.

  1. NeurIPS 2026

    MC2Mark: Distortion-Free Multi-Bit Watermarking for Long Messages

    Xuehao Cui*, Ruibo Chen*, Yihan Wu, Heng Huang

    Conference on Neural Information Processing Systems (NeurIPS), 2026 · Paper

  2. ECCV 2026

    POET: Preference Optimization for Enhanced Text-to-Image Generation

    Ruibo Chen*, Jiacheng Pan*, Heng Huang, Zhenheng Yang

    European Conference on Computer Vision (ECCV), 2026 · Paper

  3. EMNLP 2026

    More Haste, Less Speed: Weaker Single-Layer Watermark Improves Distortion-Free Watermark Ensembles

    Ruibo Chen, Yihan Wu, Xuehao Cui, Jingqi Zhang, Heng Huang

    EMNLP Main Conference, 2026 · Paper

  4. ICLR 2026

    An Ensemble Framework for Unbiased Language Model Watermarking

    Yihan Wu*, Ruibo Chen*, Georgios Milis, Heng Huang

    International Conference on Learning Representations (ICLR), 2026 · Paper

  5. NeurIPS 2025

    Robust Distortion-Free Watermark for Autoregressive Audio Generation Models

    Yihan Wu*, Georgios Milis*, Ruibo Chen*, Heng Huang

    Conference on Neural Information Processing Systems (NeurIPS), 2025 · Paper

  6. ACL 2025

    Improved Unbiased Watermark for Large Language Models

    Ruibo Chen*, Yihan Wu*, Junfeng Guo, Heng Huang

    ACL Main Conference, 2025 · Paper

  7. ICML 2025

    De-mark: Watermark Removal in Large Language Models

    Ruibo Chen*, Yihan Wu*, Junfeng Guo, Heng Huang

    International Conference on Machine Learning (ICML), 2025 · Paper

  8. ICLR 2025

    A Watermark for Order-Agnostic Language Models

    Ruibo Chen*, Yihan Wu*, Yanshuo Chen, Chenxi Liu, Junfeng Guo, Heng Huang

    International Conference on Learning Representations (ICLR), 2025 · Paper

  9. ACL 2024

    Your Vision-Language Model Itself Is a Strong Filter: Towards High-Quality Instruction Tuning with Data Selection

    Ruibo Chen, Yihan Wu, Lichang Chen, Guodong Liu, Qi He, Tianyi Xiong, Chenxi Liu, Junfeng Guo, Heng Huang

    ACL Findings, 2024 · Paper

Preprints

  1. Preprint 2026

    OriginMark: Mitigating Attacks in Distortion-Free Watermarks

    Ruibo Chen, Donghang Lu, Heng Huang, Jian Du

    Under review, 2026

  2. Preprint 2026

    TTMark: Pairwise Distortion-Free Watermarking Beyond Single-Token Entropy

    Ruibo Chen, Zhengmian Hu, Donghang Lu, Xuehao Cui, Georgios Milis, Yihan Wu, Jian Du, Heng Huang

    Under review, 2026 · Paper

  3. Preprint 2026

    Scale as a Modality: Multi-Scale Pixel-Space Diffusion Transformers

    Ruibo Chen, Kaishen Wang, Tianyi Xiong, Chenxi Liu, Yanshuo Chen, Heng Huang, Jian Du

    Under review, 2026

  4. arXiv 2025

    ImAgent: A Unified Multimodal Agent Framework for Test-Time Scalable Image Generation

    Kaishen Wang*, Ruibo Chen*, Tong Zheng, Heng Huang

    arXiv preprint, 2025 · Paper

See all publications →  ·  Google Scholar

Experience

  • 2026 Research Scientist Intern TikTok, San Jose, CA · Mentor: Jian Du
    • Developed the core watermarking algorithms for ByteDance LLM products, improving the detectability and security of distortion-free LLM watermarks.
    • Built and pretrained VAE-free multi-scale pixel-space diffusion transformers for class-conditional and text-to-image generation.
  • 2025 Research Scientist Intern TikTok, San Jose, CA · Mentors: Jiacheng Pan, Zhenheng Yang
    • Mitigated the training-inference mismatch in text-to-image models by training an LLM with RL to align user prompts with the model's input distribution, improving generation quality.

Education

  • 2023 – present Ph.D., Computer Science University of Maryland, College Park
  • 2019 – 2023 B.S., Intelligence Science and Technology Peking University

Honors & Awards

  • Dean's Fellowship, University of Maryland, College Park 2023–2025
  • Outstanding Graduates of Beijing 2023
  • Outstanding Graduates of Peking University 2023
  • National Scholarship, Ministry of Education, China 2022