AI Paper Atlas

Research map / Bandits and online learning / Contextual bandits and Thompson sampling

The Pareto-optimal Trade-off between Regret and Statistical Inference in Linear Stochastic Bandits under Safety Constraints

Yuming Shao, Zhixuan Fang

ICML 2026 · Bandits and online learning › Contextual bandits and Thompson sampling · approach: Linear and generalized linear bandits

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
lengthen Phase I (estimation) duration T1 from 70 to 700 (ablation)SERMiSC · synthetic linear bandit (T=1000, K=10, d=8)estimation errorImproved
lengthen Phase I (estimation) duration T1 from 70 to 700 (ablation)SERMiSC · synthetic linear bandit (T=1000, K=10, d=8)cumulative regret and safety violationHurt
use two-phase SERMiSC instead of PO (pessimistic-optimistic) algorithmsafe linear bandit · synthetic linear bandit (T=1000, K=10, d=8)estimation errorImproved
use SERMiSC instead of POsafe linear bandit · synthetic linear bandit (T=1000, K=10, d=8)regret and safety violationNo clear effect
ignore safety (OFUL) instead of safe explorationlinear bandit · synthetic linear bandit (T=1000, K=10, d=8)safety violationHurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy