Ying Feng

me.jpeg

Hi!

I am a third-year PhD student at MIT, where I am fortunate to be advised by Piotr Indyk. Before that, I was an undergrad at CMU.

My research interests are efficient LLM inference and high-dimensional retrieval, including KV-cache compression, quantization, approximate nearest neighbor search, and randomized data structures.

Email: yingggfeng(at)gmail.com


Manuscripts


  1. Attention under Bounded Key-Query Similarity: Space Complexity and Transformer Anisotropy
  2. Ultra-Fast Deterministic Approximate Near Neighbor Search in High Dimensions
  3. Condition-Number-Independent Sparse Linear Regression on Random Supports

Publications

2027

  1. Towards Tight Bounds for Streaming Attention
    To appear in SODA 2027

2026

  1. Provable Quantization with Randomized Hadamard Transform
    To appear in NeurIPS 2026
  2. Fast Approximate Lp Chamfer Distance via Lopsided Embeddings and Structured JL
    To appear in NeurIPS 2026
  3. Fast and Compact Random Mappings with Uniform Guarantees and Applications
    In STOC 2026

2025

  1. On Differential Privacy for Adaptively Solving Search Problems via Sketching
    In ICML 2025 (Selected for Oral Presentation)
  2. Even Faster Algorithm for the Chamfer Distance
    In ICALP 2025

2024

  1. Fast White-Box Adversarial Streaming Without a Random Oracle
    In ICML 2024
  2. A Real-Time Rescheduling Algorithm for Multi-Robot Plan Execution
    In ICAPS 2024

2023

  1. Improved Algorithms for White-Box Adversarial Streams
    In ICML 2023
  2. A Fast Rescheduling Algorithm for Real-Time Multi-Robot Coordination (Extended Abstract)
    In SoCS 2023