Tsai-chuan Wu (吳才銓)

[about] [projects] [github] [scholar] [linkedin] [cv]

About Me

I’m a member of technical staff at AMD living in San Francisco. Previously, I was a researcher at Together AI and intern at Cohere and AWS. My MSc at the University of Toronto (UofT)/Vector Institute was advised by Prof. Vardan Papyan. I completed my BSc also at UofT (Victoria College).

Recent Publications | [all]

  1. Kitty: Accurate and Efficient 2-bit KV Cache Quantization with Dynamic Channel-wise Precision Boost
    Haojun Xia, Xiaoxia Wu, Jisen Li, Robert Wu, Junxiong Wang, Jue Wang, Chenxi Li, Aman Singhal, Alay Dilipbhai Shah, Alpay Ariyak, Donglin Zhuang, Zhongzhu Zhou, Ben Athiwaratkun, Zhen Zheng, Shuaiwen Leon Song
    MLSys 2026 (Research Track)
    [arxiv]
  2. SEE-2-SOUND: Zero-Shot Spatial Environment-to-Spatial Sound
    Rishit Dagli, Shivesh Prakash, Robert Wu, Houman Khosravani
    SIGGRAPH 2025, FM-Wild @ ICML 2024
    [acm] [openreview] [arxiv] [web] [code]
  3. Linguistic Collapse: Neural Collapse in (Large) Language Models
    Robert Wu, Vardan Papyan
    NeurIPS 2024 (Main Track)
    [proceedings] [arxiv] [code]
Pre-Prints
  1. Opportunistic Expert Activation: Batch-Aware Expert Routing for Faster Decode Without Retraining
    Costin-Andrei Oncescu, Qingyang Wu, Wai Tong Chung, Robert Wu, Bryan Gopal, Junxiong Wang, Tri Dao, Ben Athiwaratkun [arxiv]
  2. Imitate Optimal Policy: Prevail and Induce Action Collapse in Policy Gradient
    Zhongzhu Zhou, Yibo Yang, Ziyan Chen, Fengxiang Bie, Haojun Xia, Xiaoxia Wu, Robert Wu, Ben Athiwaratkun, Bernard Ghanem, Shuaiwen Leon Song [arxiv]

(* equal contribution)