Robert Wu (吳才銓)

University of Toronto (UofT) → together.ai
Email: (see CV)
Links: { github, scholar, linkedin, }

[about] [projects] [cv]


About Me

I’m a junior researcher and recent MSc graduate from the University of Toronto (UofT)/Vector Institute advised by Prof. Vardan Papyan. I am generally interested in deep learning and computer systems. I completed my BSc also at UofT (Victoria College). I will be joining together.ai in January 2025.


Publications | [all]

  1. Linguistic Collapse: Neural Collapse in (Large) Language Models
    Robert Wu, Vardan Papyan
    NeurIPS 2024 (Main Track)
    [proceedings] [arxiv] [code]
  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. Towards One Shot Search Space Poisoning in Neural Architecture Search
    Nayan Saxena, Robert Wu, Rohan Jain
    AAAI 2022 (Student Abstract/Poster)
    [proceedings] [arxiv] [code]
  4. NeuralArTS: Structuring Neural Architecture Search with Type Theory
    Robert Wu, Nayan Saxena, Rohan Jain
    AAAI 2022 (Student Abstract/Poster) (top 20, oral)
    [proceedings] [arxiv] [code]
  5. Poisoning the Search Space in Neural Architecture Search
    Robert Wu*, Nayan Saxena*, Rohan Jain*
    AdvML @ ICML 2021
    [openreview] [poster] [arxiv] [code]
Pre-Prints
  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 [arxiv]
  2. 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]
  3. 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)