6 papers
Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning
Atoosa Chegini, Soheil Feizi
Selecting the best response from multiple small-model samples using a stronger scorer is a simple inference-time strategy, but fails when the small model has already committed to i…
A Single Neuron Is Sufficient to Bypass Safety Alignment in Large Language Models
Hamid Kazemi, Atoosa Chegini, Maria Safi
Safety alignment in language models operates through two mechanistically distinct systems: refusal neurons that gate whether harmful knowledge is expressed, and concept neurons tha…
Reasoning's Razor: Reasoning Improves Accuracy but Can Hurt Recall at Critical Operating Points in Safety and Hallucination Detection
Atoosa Chegini, Hamid Kazemi, Garrett Souza +5
Reasoning has become a central paradigm for large language models (LLMs), consistently boosting accuracy across diverse benchmarks. Yet its suitability for precision-sensitive task…
RePanda: Pandas-powered Tabular Verification and Reasoning
Atoosa Malemir Chegini, Keivan Rezaei, Hamid Eghbalzadeh +1
Fact-checking tabular data is essential for ensuring the accuracy of structured information. However, existing methods often rely on black-box models with opaque reasoning. We intr…
SALSA: Soup-based Alignment Learning for Stronger Adaptation in RLHF
Atoosa Chegini, Hamid Kazemi, Iman Mirzadeh +5
In Large Language Model (LLM) development, Reinforcement Learning from Human Feedback (RLHF) is crucial for aligning models with human values and preferences. RLHF traditionally re…
Rethinking Artistic Copyright Infringements in the Era of Text-to-Image Generative Models
Mazda Moayeri, Samyadeep Basu, Sriram Balasubramanian +4
Recent text-to-image generative models such as Stable Diffusion are extremely adept at mimicking and generating copyrighted content, raising concerns amongst artists that their uni…