ICML 2025 accepted papers on Bandits and online learning
96 of the 3,342 papers accepted at ICML 2025 (2.9%) are on Bandits and online learning, compared with 3.0% at ICML 2024. They fall into 4 topics.
See ICML 2025 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
Online convex optimization and games · 37 papers
- High-Dimensional Prediction for Sequential Decision Making (1 citation)
- Learning Safe Strategies for Value Maximizing Buyers in Uniform Price Auctions (1 citation)
- Near-Optimal Consistency-Robustness Trade-Offs for Learning-Augmented Online Knapsack Problems (1 citation)
- Transfer Learning for Nonparametric Contextual Dynamic Pricing (1 citation)
- Adversarial Combinatorial Semi-bandits with Graph Feedback
- An Instrumental Value for Data Production and its Application to Data Pricing
- Avoiding Catastrophe in Online Learning by Asking for Help
- Best of Both Worlds: Regret Minimization versus Minimax Play
- Beyond Minimax Rates in Group Distributionally Robust Optimization via a Novel Notion of Sparsity
- Competing Bandits in Matching Markets via Super Stability
- Exploiting Curvature in Online Convex Optimization with Delayed Feedback
- Faster Rates for Private Adversarial Bandits
and 25 more in the interactive map.
Contextual bandits and Thompson sampling · 25 papers
- A Classification View on Meta Learning Bandits
- A Parametric Contextual Online Learning Theory of Brokerage
- Catoni Contextual Bandits are Robust to Heavy-tailed Rewards
- Clustering Items through Bandit Feedback: Finding the Right Feature out of Many
- Combinatorial Reinforcement Learning with Preference Feedback
- Connecting Thompson Sampling and UCB: Towards More Efficient Trade-offs Between Privacy and Regret
- Contextual Bandits for Unbounded Context Distributions
- Contextual Linear Bandits with Delay as Payoff
- Contextual Online Decision Making with Infinite-Dimensional Functional Regression
- High Probability Bound for Cross-Learning Contextual Bandits with Unknown Context Distributions
- Improved Online Confidence Bounds for Multinomial Logistic Bandits
- Improved Regret Analysis in Gaussian Process Bandits: Optimality for Noiseless Reward, RKHS norm, and Non-Stationary Variance
and 13 more in the interactive map.
RL theory and MDPs · 17 papers
- Quantum Speedups in Regret Analysis of Infinite Horizon Average-Reward Markov Decision Processes (1 citation)
- An Optimistic Algorithm for online CMDPS with Anytime Adversarial Constraints
- Automatic Reward Shaping from Confounded Offline Data
- Bellman Unbiasedness: Toward Provably Efficient Distributional Reinforcement Learning with General Value Function Approximation
- Concurrent Reinforcement Learning with Aggregated States via Randomized Least Squares Value Iteration
- Gap-Dependent Bounds for Federated $Q$-Learning
- Learning Adversarial MDPs with Stochastic Hard Constraints
- Logarithmic Regret for Online KL-Regularized Reinforcement Learning
- Near-optimal Regret Using Policy Optimization in Online MDPs with Aggregate Bandit Feedback
- Online Episodic Convex Reinforcement Learning
- Minimax Optimal Regret Bound for Reinforcement Learning with Trajectory Feedback
- Multi-objective Linear Reinforcement Learning with Lexicographic Rewards
and 5 more in the interactive map.
Multi-armed bandits · 17 papers
- BanditSpec: Adaptive Speculative Decoding via Bandit Algorithms
- Constrained Pareto Set Identification with Bandit Feedback
- Fixed-Confidence Multiple Change Point Identification under Bandit Feedback
- Fixing the Loose Brake: Exponential-Tailed Stopping Time in Best Arm Identification
- Fusing Reward and Dueling Feedback in Stochastic Bandits
- Near Optimal Best Arm Identification for Clustered Bandits
- Near Optimal Non-asymptotic Sample Complexity of 1-Identification
- Offline Learning for Combinatorial Multi-armed Bandits
- The Batch Complexity of Bandit Pure Exploration
- Tokenized Bandit for LLM Decoding and Alignment
- Tracking Most Significant Shifts in Infinite-Armed Bandits
- Adaptive Data Collection for Robust Learning Across Multiple Distributions
and 5 more in the interactive map.