Lune

Profiles2020-2025

Ashok Cutkosky

Boston University Boston

29Papers
7Last author
25Unique co-authors
8First author
3Solo

First and last author come from byline position, which some venues assign alphabetically.

Recent papers29

  1. Descent with Misaligned Gradients and Applications to Hidden ConvexityAuthor 2 of 4ICLR 2025
  2. General framework for online-to-nonconvex conversion: Schedule-free SGD is also effective for nonconvex optimizationAuthor 4 of 4ICML 2025
  3. Unconstrained Robust Online Convex OptimizationAuthor 2 of 2ICML 2025
  4. Adam with model exponential moving average is effective for nonconvex optimizationAuthor 2 of 2NeurIPS 2024 · 36 citations
  5. Fully Unconstrained Online LearningAuthor 1 of 2NeurIPS 2024 · 13 citations
  6. Optimal Stochastic Non-smooth Non-convex Optimization through Online-to-Non-convex ConversionAuthor 1 of 3ICML 2023 · 54 citations
  7. Long Range Language Modeling via Gated State SpacesAuthor 3 of 4ICLR 2023 · 45 citations
  8. Mechanic: A Learning Rate TunerAuthor 1 of 3NeurIPS 2023 · 27 citations
  9. Unconstrained Dynamic Regret via Sparse CodingAuthor 2 of 4NeurIPS 2023 · 14 citations
  10. Alternation makes the adversary weaker in two-player gamesAuthor 2 of 5NeurIPS 2023 · 8 citations
  11. Bandit Online Linear Optimization with Hints and QueriesAuthor 2 of 4ICML 2023 · 4 citations
  12. Parameter-free Regret in High Probability with Heavy TailsAuthor 2 of 2NeurIPS 2022 · 41 citations
  13. PDE-Based Optimal Strategy for Unconstrained Online LearningAuthor 2 of 4ICML 2022 · 31 citations
  14. Better SGD using Second-order MomentumAuthor 2 of 2NeurIPS 2022 · 18 citations
  15. Momentum Aggregation for Private Non-convex ERMAuthor 2 of 2NeurIPS 2022 · 14 citations
  16. Optimal Comparator Adaptive Online Learning with Switching CostAuthor 2 of 4NeurIPS 2022 · 10 citations
  17. Differentially Private Online-to-batch for Smooth LossesAuthor 3 of 3NeurIPS 2022 · 5 citations
  18. High-probability Bounds for Non-Convex Stochastic Optimization with Heavy TailsAuthor 1 of 2NeurIPS 2021 · 119 citations
  19. Dynamic Balancing for Model Selection in Bandits and RLAuthor 1 of 6ICML 2021 · 40 citations
  20. Robust Pure Exploration in Linear Bandits with Limited BudgetAuthor 2 of 3ICML 2021 · 27 citations
  21. Logarithmic Regret from Sublinear HintsAuthor 2 of 4NeurIPS 2021 · 23 citations
  22. Extreme Memorization via Scale of InitializationAuthor 2 of 3ICLR 2021 · 22 citations
  23. Online Selective Classification with Limited FeedbackAuthor 3 of 4NeurIPS 2021 · 12 citations
  24. Momentum Improves Normalized SGDAuthor 1 of 1ICML 2020 · 177 citations
  25. Online Learning with Imperfect HintsAuthor 2 of 4ICML 2020 · 64 citations
  26. Parameter-free, Dynamic, and Strongly-Adaptive Online LearningAuthor 1 of 1ICML 2020 · 63 citations
  27. Online Linear Optimization with Many HintsAuthor 3 of 4NeurIPS 2020 · 21 citations
  28. Comparator-Adaptive Convex BanditsAuthor 2 of 3NeurIPS 2020 · 14 citations
  29. Better Full-Matrix Regret via Parameter-Free Online LearningAuthor 1 of 1NeurIPS 2020 · 7 citations
Dusk over the sea between two cliffs drawn in fine vertical lines