papers

Selected papers, preprints, publications, and miscellaneous writing.

(α-β) indicates alphabetical author ordering, and * indicates equal contribution.

Selected papers

  1. Operator Learning: A Statistical Perspective
    Unique Subedi and Ambuj Tewari
    Annual Review of Statistics and Its Applications, 2026
  2. On the Benefits of Active Data Collection in Operator Learning (Spotlight)
    Unique Subedi and Ambuj Tewari
    International Conference on Machine Learning, 2025
  3. A Unified Theory of Supervised Online Learnability (Outstanding Paper Award)
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Conference on Algorithmic Learning Theory, 2025

Preprints

  1. Operator Learning for Schrödinger Equation: Unitarity, Error Bounds, and Time Generalization
    Yash Patel*, Unique Subedi*, and Ambuj Tewari
    Preprint, 2025

Publications

2026

  1. Learning to Choose or Choosing to Learn: Best-of-N vs. Supervised Fine-Tuning for Bit String Generation
    Seamus Somerstep, Vinod Raman, Unique Subedi, and 1 more author
    International Conference on Artificial Intelligence and Statistics, 2026
  2. Operator Learning: A Statistical Perspective
    Unique Subedi and Ambuj Tewari
    Annual Review of Statistics and Its Applications, 2026

2025

  1. Controlling Statistical, Discretization, and Truncation Errors in Learning Fourier Linear Operators
    Unique Subedi and Ambuj Tewari
    Transactions on Machine Learning Research, 2025
  2. On the Benefits of Active Data Collection in Operator Learning (Spotlight)
    Unique Subedi and Ambuj Tewari
    International Conference on Machine Learning, 2025
  3. The Complexity of Sequential Prediction in Dynamical Systems (Oral)
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Learning for Dynamics and Control Conference, 2025
  4. A Unified Theory of Supervised Online Learnability (Outstanding Paper Award)
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Conference on Algorithmic Learning Theory, 2025

2024

  1. A Characterization of Multioutput Learnability
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Journal of Machine Learning Research, 2024
  2. Multiclass Transductive Online Learning (Spotlight)
    (α-β) Steve Hanneke, Vinod Raman, Amirreza Shaeiri, and 1 more author
    Advances in Neural Information Processing Systems, 2024
  3. Smoothed Online Classification can be Harder than Batch Classification
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Advances in Neural Information Processing Systems, 2024
  4. Online Learning with Set-Valued Feedback
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Conference on Learning Theory, 2024
  5. Apple Tasting: Combinatorial Dimensions and Minimax Rates
    Vinod Raman*, Unique Subedi*, Ananth Raman, and 1 more author
    Conference on Learning Theory, 2024
  6. Online Infinite-Dimensional Regression: Learning Linear Operators
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Conference on Algorithmic Learning Theory, 2024
  7. Multiclass Online Learnability under Bandit Feedback
    Ananth Raman, Vinod Raman*, Unique Subedi*, and 2 more authors
    Conference on Algorithmic Learning Theory, 2024

2023

  1. On the Learnability of Multilabel Ranking (Spotlight)
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Advances in Neural Information Processing Systems, 2023
  2. On Proper Learnability between Average- and Worst-case Robustness
    Vinod Raman*, Unique Subedi*, and Ambuj Tewari
    Advances in Neural Information Processing Systems, 2023
  3. Multiclass Online Learning and Uniform Convergence
    (α-β) Steve Hanneke, Shay Moran, Vinod Raman, and 2 more authors
    Conference on Learning Theory, 2023

2022

  1. A Weighted Version of Erdős-Kac Theorem
    (α-β) Rizwanur Khan, Micah B. Milinovich, and Unique Subedi
    Journal of Number Theory, 2022
  2. A Conjectural Asymptotic Formula for Multiplicative Chaos in Number Theory
    (α-β) Daksh Aggarwal, Unique Subedi, William Verreault, and 2 more authors
    Research in Number Theory, 2022
  3. Sums of Random Multiplicative Functions over Function Fields with Few Irreducible Factors
    (α-β) Daksh Aggarwal, Unique Subedi, William Verreault, and 2 more authors
    Mathematical Proceedings of the Cambridge Philosophical Society, 2022

Thesis

  1. Learning Theory in the AI for Science Era: From Classical Foundations to Operator Learning
    Unique Subedi
    University of Michigan, 2026
    PhD thesis
  2. A Weighted Version of Erdős-Kac Theorem
    Unique Subedi
    University of Mississippi, 2022
    Undergraduate thesis

Miscellaneous

  1. riemann-zeta.jpg
    An Elementary Evaluation of ζ(2n) Using Dirichlet’s Kernel
    Unique Subedi and Micah B. Milinovich
    2022
    Technical Note
  2. dirichlet-kernel.png
    On L_1-norm of Dirichlet’s Kernel
    Unique Subedi and Micah B. Milinovich
    2022
    Technical Note