For linear bandits with actions in the Euclidean unit ball, Bayesian incentive-compatible exploration is achievable with polynomial sample complexity, eliminating the exponential barriers known for polytope action sets.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.GT 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Geometry Meets Incentives: Sample-Efficient Incentivized Exploration with Linear Contexts
For linear bandits with actions in the Euclidean unit ball, Bayesian incentive-compatible exploration is achievable with polynomial sample complexity, eliminating the exponential barriers known for polytope action sets.