Curriculum vitae

Brian K Chen

Independent Researcher

e0694208@u.nus.edu

Research interests

Model efficiency and reasoning; context, memory, and expressivity in transformers; diffusion language models; efficient inference and adaptation.

Education

National University of Singapore

2021–2025

PhD in Computer Science

Supervisor: Kenji Kawaguchi

University of Oxford

2016–2020

MMath in Mathematics & Statistics

Experience

Independent Researcher

Current

Research on model efficiency and reasoning, context and memory in transformers, and diffusion language models.

Esquel Group

2019

HR Data Analyst Intern

Developed an R-based model for three-month employee attrition prediction.

Selected publications

* Equal contribution

  1. PrefixMemory-Tuning: Modernizing Prefix-Tuning by Decoupling the Prefix from Attention

    Haonan Wang*, Brian K Chen*, Siquan Li*, Xinhe Liang, Hwee Kuan Lee, Kenji Kawaguchi, Tianyang Hu

    ICLR 2026 · Co-first author

  2. Rethinking Pathology Image Analysis through Shuffling

    Zeyu Liu*, Tianyi Zhang*, Brian K Chen*, et al.

    npj Biomedical Innovations, 2026 · Co-first author

  3. Exact Conversion of In-Context Learning to Model Weights in Linearized-Attention Transformers

    Brian K Chen, Tianyang Hu, Hui Jin, Hwee Kuan Lee, Kenji Kawaguchi

    ICML 2024 · First author

  4. In-Context Learning Behaves as a Greedy Layer-Wise Gradient Descent Algorithm

    Brian K Chen, Tianyang Hu, Hui Jin, Hwee Kuan Lee, Kenji Kawaguchi

    NeurIPS Workshop, 2024 · First author

Submitted manuscripts

ICLR 2027 submissions

  1. The Value of Memory in Language Models: Joint Limits on Language Model Expressivity

    Brian K Chen, Cui Kai

  2. How Far Can Sliding Window Attention See?

    Maxime Meyer*, Brian K Chen*, Nikolaj Murašov, Caroline Chaux, Vincent Y. F. Tan

  3. Reasoning Under Intervention: Learned Soft Prefixes as Behavioral Probes

    Brian K Chen

  4. Speculative Correction: Draft-Then-Refine Decoding for Diffusion Language Models

    Brian K Chen, Chong Wu, Kenji Kawaguchi

Work in progress

Skills & languages

Programming: Python, R
Languages: English (native), Mandarin (native)