Dictionary Learning Improves Patch-Free Circuit Discovery in Mechanistic Interpretability: A Case Study on Othello-GPTby Zhengfu…
Reformatted Alignmentby Run-Ze Fan, Xuefeng Li, Haoyang Zou, Junlong Li, Shwai He, Ethan Chern, Jiewen…
Diffusion Tempering Improves Parameter Estimation with Probabilistic Integrators for Ordinary Differential Equationsby Jonas Beck, Nathanael…
AnyGPT: Unified Multimodal LLM with Discrete Sequence Modelingby Jun Zhan, Junqi Dai, Jiasheng Ye, Yunhua…
Kernel KMeans clustering splits for end-to-end unsupervised decision treesby Louis Ohl, Pierre-Alexandre Mattei, Mickaël Leclercq,…
The Fundamental Limits of Least-Privilege Learningby Theresa Stadler, Bogdan Kulynych, Michael C. Gastpar, Nicolas Papernot,…
Convergence of Gradient Descent for Recurrent Neural Networks: A Nonasymptotic Analysisby Semih Cayci, Atilla EryilmazFirst…
BEARS Make Neuro-Symbolic Models Aware of their Reasoning Shortcutsby Emanuele Marconato, Samuele Bortolotti, Emile van…
Learning to Defer in Content Moderation: The Human-AI Interplayby Thodoris Lykouris, Wentao WengFirst submitted to…
Synthetic location trajectory generation using categorical diffusion modelsby Simon Dirmeier, Ye Hong, Fernando Perez-CruzFirst submitted…