Francesco Orabona

Nonparametric Teaching of Attention Learners

A Finite-Time Analysis of TD Learning with Linear Function Approximation without Projections or Strong Convexity

Dynamic Regret Reduces to Kernelized Static Regret

A Unified Theoretical Analysis of Private and Robust Offline Alignment: from RLHF to DPO

New Perspectives on the Polyak Stepsize: Surrogate Functions and Negative Results

Optimal Regret of Bandits under Differential Privacy

Square$\chi$PO: Differentially Private and Robust $\chi^2$-Preference Optimization in Offline Direct Alignment

ATA: Adaptive Task Allocation for Efficient Resource Management in Distributed Machine Learning

STAR-Bets: Sequential TArget-Recalculating Bets for Tighter Confidence Intervals