mlsanityData to 14 Oct 2026

Bandit Guided Submodular Curriculum for Adaptive Subset Selection

arXiv 2511.22944 ↗First posted 28 Nov 2025cs.LG cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

Traditional curriculum learning proceeds from easy to hard samples, yet defining a reliable notion of difficulty remains elusive. Prior work has used submodular functions to induce difficulty scores in curriculum learning. We reinterpret adaptive subset selection and formulate it as a multi-armed bandit problem, where each arm corresponds to a submodular function guiding sample selection. We introduce ONLINESUBMOD, a novel online greedy policy that optimizes a utility-driven reward and provably achieves no-regret performance under various sampling regimes. Empirically, ONLINESUBMOD outperforms both traditional curriculum learning and bi-level optimization approaches across vision and language datasets, showing superior accuracy-efficiency tradeoffs. More broadly, we show that validationdriven reward metrics offer a principled way to guide the curriculum schedule.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 37 of its 58 references that are in the map. First posted on arXiv 28 Nov 2025.

Builds on

37 of 58 references.

References · 37

  1. 2015Deep Residual Learning for Image Recognitioncited by 52,096
  2. 2019Sentence-BERT: Sentence Embeddings using Siamese BERT-Networkscited by 9,170
  3. 2017Gradient Episodic Memory for Continual Learningcited by 1,898
  4. 2018On First-Order Meta-Learning Algorithmscited by 1,018
  5. 2015Optimizing Neural Networks with Kronecker-factored Approximate Curvaturecited by 796
  6. 2019Three scenarios for continual learningcited by 614
  7. 2019Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Boundscited by 610
  8. 2020Estimating Training Data Influence by Tracing Gradient Descentcited by 580
  9. 2024LESS: Selecting Influential Data for Targeted Instruction Tuningcited by 532
  10. 2020Curriculum Learning for Reinforcement Learning Domains: A Framework and Surveycited by 373
  11. 2018Not All Samples Are Created Equal: Deep Learning with Importance Samplingcited by 328
  12. 2017Empirical Analysis of the Hessian of Over-Parametrized Neural Networkscited by 290
25 more
  1. 2021GRAD-MATCH: Gradient Matching based Data Subset Selection for Efficient Deep Model Trainingcited by 274
  2. 2022Prioritized Training on Points that are Learnable, Worth Learning, and Not Yet Learntcited by 236
  3. 2024Continual Learning of Large Language Models: A Comprehensive Surveycited by 234
  4. 2020GLISTER: Generalization based Data Subset Selection for Efficient and Robust Learningcited by 226
  5. 2018Task-Free Continual Learningcited by 221
  6. 2015Online Batch Selection for Faster Training of Neural Networkscited by 122
  7. 2022Coverage-centric Coreset Selection for High Pruning Ratescited by 117
  8. 2023D2 Pruning: Message Passing for Balancing Diversity and Difficulty in Data Pruningcited by 69
  9. 2016Restricted Strong Convexity Implies Weak Submodularitycited by 52
  10. 2021PRISM: A Rich Class of Parameterized Submodular Information Measures for Guided Subset Selectioncited by 51
  11. 2020Dissecting Hessian: Understanding Common Structure of Hessian in Neural Networkscited by 45
  12. 2017Guarantees for Greedy Maximization of Non-submodular Functions with Applicationscited by 35
  13. 2017Gradient Methods for Submodular Maximizationcited by 33
  14. 2022Prioritizing Samples in Reinforcement Learning with Reducible Losscited by 22
  15. 2024Curriculum Learning for Small Code Language Modelscited by 22
  16. 2020Sampling from a k-DPP without looking at all itemscited by 17
  17. 2025Data Pruning by Information Maximizationcited by 15
  18. 2018Distributionally Robust Submodular Maximizationcited by 12
  19. 2023Curriculum Learning for Graph Neural Networks: Which Edges Should We Learn Firstcited by 10
  20. 2023MILO: Model-Agnostic Subset Selection Framework for Efficient Model Training and Tuningcited by 9
  21. 2023Nearly Minimax Optimal Submodular Maximization with Bandit Feedbackcited by 7
  22. 2021Convergence Guarantees for Deep Epsilon Greedy Policy Learningcited by 6
  23. 2018An Optimal Algorithm for Online Unconstrained Submodular Maximizationcited by 3
  24. 2025Online Nonsubmodular Optimization with Delayed Feedback in the Bandit Settingcited by 1
  25. 2025Bayesian Coreset Optimization for Personalized Federated Learningcited by 1

Built on by

Not cited yet.