Risk-Sensitive Non-Parametric Bandits Match Classical Regret Bounds
Non-parametric CPT bandits incur no additional order-wise regret cost compared to standard bandits under Cumulative Prospect Theory.
Stories per day · 30 days
1 RSS
Relatedblack-box-optimization1gaussian-process1matrn1monte-carlo-tree-search1regret1reinforcement-learning1resource-allocation1risk-sensitive1scientific-discovery1
10 stories, newest first · one card per story
Stories tagged #bandits
Gaussian process bandits proposed for constrained black-box lever optimization
Authors report a statistically significant efficiency gain versus epsilon-greedy in one live production randomized trial.
Paper defines recommender regret for regularised RL estimates
Bandit and MCTS estimators are recast as recommenders; experiments cover synthetic trees, arcade games, and scientific tasks.
Best Arm Identification for Bandits with Shifting Means
ISM identifies the best arm when shared mean shifts are adversarial but gaps stay fixed.
On the Computational Tractability of Robust Bandits
Researchers give a polynomial-time robust bandit learner with near-square-root regret and show nearby variants are NP-hard.
When May a Bandit Leave Its Anchor? E-Process-Authorized Thompson Sampling under Non-stationarity
Paper proposes e-process-authorized Thompson sampling (e-ATS), using anytime-valid e-processes to decide when bandits may forget history under non-stationary rewards.
Minimax Optimal Regret for Causal Logistic Bandits with Counterfactual Fairness
The paper proves matching minimax regret bounds for causal logistic bandits under counterfactual fairness.
ActiveSaddler: Automated Curriculum Learning for Agent Harness Optimization
ActiveSaddler adapts agent-harness curricula, raising Pass@1 by 4.4 and 7.5 points on two benchmarks.
A positive resolution of the gap-entropy conjecture
New proof resolves the gap-entropy conjecture for Gaussian bandits, bounding optimal best-arm identification samples by H(log(1/delta)+Ent(I)) up to constants.
Online Learning with LLM Experts from Limited Feedback
Paper proposes bandit algorithms for adaptively routing prompts to LLM experts, minimizing regret under limited feedback budgets.