collaborators

10 papers

cs.AI2026

On the Generalization of Steering Vectors for Chain-of-Thought Faithfulness

Matthew Nguyen, Kyle Cox, Austin Meek +1

Model capabilities have improved in large part due to scaling chain of thought. This has been a promising development for AI safety--where models verbalize their reasoning, it is p…

cs.AI2026

Base Models Know How to Reason, Thinking Models Learn When

Constantin Venhoff, Iván Arcuschin, Philip Torr +2

What do thinking language models learn during training that their base models lack? We first present an unsupervised method that discovers a model's reasoning behaviors by training…

cs.AI2026

Chain-of-Thought Reasoning In The Wild Is Not Always Faithful

Iván Arcuschin, Jett Janiak, Robert Krzyzanowski +3

Recent studies indicate that when faced with explicit biases in prompts, models often omit mentioning these biases in their Chain-of-Thought (CoT) output, revealing that verbalized…

cs.LG2026

Biases in the Blind Spot: Detecting What LLMs Fail to Mention

Iván Arcuschin, David Chanin, Adrià Garriga-Alonso +1

Large Language Models (LLMs) often provide chain-of-thought (CoT) reasoning traces that appear plausible, but may hide internal biases. We call these unverbalized biases. Monitorin…

cs.LG2026

Automatically Finding Reward Model Biases

Atticus Wang, Iván Arcuschin, Arthur Conmy

Reward models are central to large language model (LLM) post-training. However, past work has shown that they can reward spurious or undesirable attributes such as length, format,…

cs.LG2026

Mind the Performance Gap: Capability-Behavior Trade-offs in Feature Steering

Eitan Sprejer, Oscar Agustín Stanchi, María Victoria Carro +2

Feature steering has emerged as a promising approach for controlling LLM behavior through direct manipulation of internal representations, offering advantages over prompt engineeri…