PPaperPicks

Zhao Song

University of California Berkeley, CA, USA

52 papers at tracked venues · 33 at CORE A* · active 2024–2026

DBLP profile ↗ORCID 0000-0003-4589-5234 ↗Google Scholar ↗Homepage ↗

Venues

NeurIPS ×11ICML ×9AISTATS ×8ICLR ×8UAI ×4AAAI ×3EMNLP ×3WACV ×2CIKM ×1ICCV ×1ICDM ×1SIGKDD ×1

Frequent coauthors

Yingyu Liang
DBLP profile ↗ORCID search ↗
×5
Yuzhou Gu
DBLP profile ↗ORCID search ↗
×4
Jerry Yao-Chieh Hu
DBLP profile ↗ORCID search ↗
×4
Yeqi Gao
DBLP profile ↗ORCID search ↗
×3
Bo Chen
DBLP profile ↗ORCID search ↗
×3
Yang Cao
DBLP profile ↗ORCID search ↗
×3
Xiaoyu Li
DBLP profile ↗ORCID search ↗
×2
Chenyang Li
DBLP profile ↗ORCID search ↗
×2
Xuan Shen
DBLP profile ↗ORCID search ↗
×2
Josh Alman
DBLP profile ↗ORCID search ↗
×2
Timothy Chu
DBLP profile ↗ORCID search ↗
×2
Yubin Chen
DBLP profile ↗ORCID search ↗
×1

Papers

  1. T2VWorldBench: A Benchmark for Evaluating World Knowledge in Text-to-Video Generation
    WACV 2026 ·
    Yubin Chen
    DBLP profile ↗ORCID search ↗
  2. A Fast Optimization View: Reformulating Single Layer Attention in LLM Based on Tensor and SVM Trick, and Solving It in Matrix Multiplication Time
    UAI 2025 ·
    Yeqi Gao
    DBLP profile ↗ORCID search ↗
  3. An Iterative Algorithm for Rescaled Hyperbolic Functions Regression
    AISTATS 2025 ·
    Yeqi Gao
    DBLP profile ↗ORCID search ↗
  4. Attention Mechanism, Max-Affine Partition, and Universal Approximation
    NeurIPS 2025 ·
    Hude Liu
    DBLP profile ↗ORCID search ↗
  5. Beyond Linear Approximations: A Novel Pruning Approach for Attention Matrix
    ICLR 2025 ·
    Yingyu Liang
    DBLP profile ↗ORCID search ↗
  6. Binary Hypothesis Testing for Softmax Models and Leverage Score Models
    ICML 2025 ·
    Yuzhou Gu
    DBLP profile ↗ORCID search ↗
  7. Bypassing the Exponential Dependency: Looped Transformers Efficiently Learn In-context by Multi-step Gradient Descent
    AISTATS 2025 ·
    Bo Chen
    DBLP profile ↗ORCID search ↗
  8. Circuit Complexity Bounds for RoPE-based Transformer Architecture
    EMNLP 2025 ·
    Bo Chen
    DBLP profile ↗ORCID search ↗
  9. Computational Limits of Low-Rank Adaptation (LoRA) Fine-Tuning for Transformer Models
    ICLR 2025 ·
    Jerry Yao-Chieh Hu
    DBLP profile ↗ORCID search ↗
  10. Conv-Basis: A New Paradigm for Efficient Attention Inference and Gradient Computation in Transformers
    EMNLP 2025 ·
    Yingyu Liang
    DBLP profile ↗ORCID search ↗
  11. Deterministic Sparse Fourier Transform for Continuous Signals with Frequency Gap
    ICML 2025 ·
    Xiaoyu Li
    DBLP profile ↗ORCID search ↗
  12. Differential Privacy Mechanisms in Neural Tangent Kernel Regression
    WACV 2025 ·
    Jiuxiang Gu
    DBLP profile ↗ORCID search ↗
  13. Differential Privacy for Euclidean Jordan Algebra with Applications to Private Symmetric Cone Programming
    NeurIPS 2025 · Zhao Song
  14. Discrepancy Minimization in Input-Sparsity Time
    ICML 2025 ·
    Yichuan Deng
    DBLP profile ↗ORCID search ↗
  15. Dissecting Submission Limit in Desk-Rejections: A Mathematical Analysis of Fairness in AI Conference Policies
    ICML 2025 ·
    Yuefan Cao
    DBLP profile ↗ORCID search ↗
  16. Dynamic Maintenance of Kernel Density Estimation Data Structure: From Practice to Theory
    UAI 2025 ·
    Jiehao Liang
    DBLP profile ↗ORCID search ↗
  17. Efficient $k$-Sparse Band-Limited Interpolation with Improved Approximation Ratio
    NeurIPS 2025 ·
    Yang Cao
    DBLP profile ↗ORCID search ↗
  18. Efficient Alternating Minimization with Applications to Weighted Low Rank Approximation
    ICLR 2025 · Zhao Song
  19. Fast Sampling for Privacy-Preserving Lazy Multiplicative Weight Update
    ICDM 2025 ·
    Xiaoyu Li
    DBLP profile ↗ORCID search ↗
  20. Faster Algorithms for Structured John Ellipsoid Computation
    NeurIPS 2025 ·
    Yang Cao
    DBLP profile ↗ORCID search ↗
  21. Faster Algorithms for Structured Linear and Kernel Support Vector Machines
    ICLR 2025 ·
    Yuzhou Gu
    DBLP profile ↗ORCID search ↗
  22. Force Matching with Relativistic Constraints: A Physics-Inspired Approach to Stable and Efficient Generative Modeling
    CIKM 2025 ·
    Yang Cao
    DBLP profile ↗ORCID search ↗
  23. Fourier Circuits in Neural Networks and Transformers: A Case Study of Modular Arithmetic with Multiple Inputs
    AISTATS 2025 ·
    Chenyang Li
    DBLP profile ↗ORCID search ↗
  24. Fundamental Limits of Prompt Tuning Transformers: Universality, Capacity and Efficiency
    ICLR 2025 ·
    Jerry Yao-Chieh Hu
    DBLP profile ↗ORCID search ↗
  25. Fundamental Limits of Visual Autoregressive Transformers: Universal Approximation Abilities
    ICML 2025 ·
    Yifang Chen
    DBLP profile ↗ORCID search ↗
  26. High-Order Flow Matching: Unified Framework and Sharp Statistical Rates
    NeurIPS 2025 ·
    Maojiang Su
    DBLP profile ↗ORCID search ↗
  27. In-Context Deep Learning via Transformer Models
    ICML 2025 ·
    Weimin Wu
    DBLP profile ↗ORCID search ↗
  28. LazyDiT: Lazy Learning for the Acceleration of Diffusion Transformers
    AAAI 2025 ·
    Xuan Shen
    DBLP profile ↗ORCID search ↗
  29. Looped ReLU MLPs May Be All You Need as Practical Programmable Computers
    AISTATS 2025 ·
    Yingyu Liang
    DBLP profile ↗ORCID search ↗
  30. NRFlow: Towards Noise-Robust Generative Modeling via High-Order Mechanism
    UAI 2025 ·
    Bo Chen
    DBLP profile ↗ORCID search ↗
  31. Numerical Pruning for Efficient Autoregressive Models
    AAAI 2025 ·
    Xuan Shen
    DBLP profile ↗ORCID search ↗
  32. On Differential Privacy for Adaptively Solving Search Problems via Sketching
    ICML 2025 ·
    Shiyuan Feng
    DBLP profile ↗ORCID search ↗
  33. The Expressibility of Polynomial based Attention Scheme
    SIGKDD 2025 · Zhao Song
  34. Towards Infinite-Long Prefix in Transformer
    EMNLP 2025 ·
    Yingyu Liang
    DBLP profile ↗ORCID search ↗
  35. Unraveling the Smoothness Properties of Diffusion Models: A Gaussian Mixture Perspective
    ICCV 2025 ·
    Yingyu Liang
    DBLP profile ↗ORCID search ↗
  36. When Can We Solve the Weighted Low Rank Approximation Problem in Truly Subquadratic Time?
    AISTATS 2025 ·
    Chenyang Li
    DBLP profile ↗ORCID search ↗
  37. A General Algorithm for Solving Rank-one Matrix Sensing
    AISTATS 2024 ·
    Lianke Qin
    DBLP profile ↗ORCID search ↗
  38. A Sublinear Adversarial Training Algorithm
    ICLR 2024 ·
    Yeqi Gao
    DBLP profile ↗ORCID search ↗
  39. Algorithm and Hardness for Dynamic Attention Maintenance in Large Language Models
    ICML 2024 ·
    Jan van den Brand
    DBLP profile ↗ORCID search ↗
  40. Fast Dynamic Sampling for Determinantal Point Processes
    AISTATS 2024 · Zhao Song
  41. How to Capture Higher-order Correlations? Generalizing Matrix Softmax Attention to Kronecker Computation
    ICLR 2024 ·
    Josh Alman
    DBLP profile ↗ORCID search ↗
  42. How to Protect Copyright Data in Optimization of Large Language Models?
    AAAI 2024 ·
    Timothy Chu
    DBLP profile ↗ORCID search ↗
  43. Log-concave Sampling from a Convex Body with a Barrier: a Robust and Unified Dikin Walk
    NeurIPS 2024 ·
    Yuzhou Gu
    DBLP profile ↗ORCID search ↗
  44. Low Rank Matrix Completion via Robust Alternating Minimization in Nearly Linear Time
    ICLR 2024 ·
    Yuzhou Gu
    DBLP profile ↗ORCID search ↗
  45. Metric Transforms and Low Rank Representations of Kernels for Fast Attention
    NeurIPS 2024 ·
    Timothy Chu
    DBLP profile ↗ORCID search ↗
  46. On Computational Limits of Modern Hopfield Models: A Fine-Grained Complexity Analysis
    ICML 2024 ·
    Jerry Yao-Chieh Hu
    DBLP profile ↗ORCID search ↗
  47. On Convergence of Federated Averaging Langevin Dynamics
    UAI 2024 ·
    Wei Deng
    DBLP profile ↗ORCID search ↗
  48. On Socially Fair Low-Rank Approximation and Column Subset Selection
    NeurIPS 2024 · Zhao Song
  49. On Statistical Rates and Provably Efficient Criteria of Latent Diffusion Transformers (DiTs)
    NeurIPS 2024 ·
    Jerry Yao-Chieh Hu
    DBLP profile ↗ORCID search ↗
  50. Solving Attention Kernel Regression Problem via Pre-conditioner
    AISTATS 2024 · Zhao Song
  51. The Closeness of In-Context Learning and Weight Shifting for Softmax Regression
    NeurIPS 2024 ·
    Shuai Li
    DBLP profile ↗ORCID search ↗
  52. The Fine-Grained Complexity of Gradient Computation for Training Large Language Models
    NeurIPS 2024 ·
    Josh Alman
    DBLP profile ↗ORCID search ↗