Wenqian Weng — accepted AI-conference papers

2026

Improved Bounds for Private and Robust Alignment

ICML 2026poster

In this paper, we study the private and robust alignment of language models from a theoretical perspective by establishing upper bounds on the suboptimality gap in both offline and online settings. We consider preference labels subject to privacy constraints and/or adversarial corruption, and analyz…

Cited by 0 Source ↗Scholar ↗

2025

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

Xingyu Zhou, Yulian Wu, Wenqian Weng, Francesco Orabona

ICML 2025poster

In this paper, we theoretically study the offline alignment of language models with human preference feedback, under both preference label corruption and privacy protections. To this end, we propose a variant of \texttt{$\chi$PO} -- \texttt{Square}\texttt{$\chi$PO}, which is a simple one-line change…

Cited by 0 Source ↗PDF ↗Scholar ↗