Ax Sofiane Ennadir, Tianze Wang, Oleg Smirnov, Sahar Asadi, Lele Cao 2/20/2026

Be Wary of Your Time Series Preprocessing

Theoretical analysis of normalization strategies and their impact on expressivity of Transformer-based time series models.

Ax Baihe Huang, Eric Xu, Kannan Ramchandran, Jiantao Jiao, Michael I. Jordan 2/20/2026

Towards Anytime-Valid Statistical Watermarking

Develops anytime-valid statistical watermarking method to distinguish machine-generated text from human content in LLMs.

Ax Payel Bhattacharjee, Osvaldo Simeone, Ravi Tandon 2/20/2026

MARS: Margin-Aware Reward-Modeling with Self-Refinement

Margin-aware reward modeling framework with self-refinement for RLHF/RLAIF alignment pipelines, reducing reliance on human preference data through augmentation.

Ax Simon Lermen, Daniel Paleka, Joshua Swanson, Michael Aerni, Nicholas Carlini, Florian Tram\`er 2/20/2026

Large-scale online deanonymization with LLMs

Large-scale deanonymization attack using LLM agents with internet access to re-identify pseudonymous online profiles.

Ax Yiqing Xie, Emmy Liu, Gaokai Zhang, Nachiket Kotalwar, Shubham Gandhi, Sathwik Acharya, Xingyao Wang, Carolyn Rose, Graham Neubig, Daniel Fried 2/20/2026

Hybrid-Gym: Training Coding Agents to Generalize Across Tasks

Hybrid-Gym environment for training coding agents on diverse software engineering tasks beyond single GitHub issues.

Ax Sasha Behrouzi, Lichao Wu, Mohamadreza Rostami, Ahmad-Reza Sadeghi 2/20/2026

NeST: Neuron Selective Tuning for LLM Safety

NeST selective neuron tuning approach for parameter-efficient LLM safety alignment without full fine-tuning overhead.

Ax Etienne Lempereur, Nathana\"el Cuvelle--Magar, Florentin Coeurdoux, St\'ephane Mallat, Eric Vanden-Eijnden 2/20/2026

MGD: Moment Guided Diffusion for Maximum Entropy Generation

Diffusion-based method for generating high-dimensional samples from moment constraints with maximum entropy guarantees.

Ax Mateusz Nowak, Xavier Cadet, Peter Chin 2/20/2026

ABCD: All Biases Come Disguised

Study of position and label biases in LLM multiple-choice question answering via synthetic benchmark evaluation.

Ax Aidar Myrzakhan, Tianyi Li, Bowei Guo, Shengkun Tang, Zhiqiang Shen 2/20/2026

Sink-Aware Pruning for Diffusion Language Models

Pruning technique for diffusion language models reducing inference cost by reconsidering attention sink preservation.