11 papers
When Policies Change Probabilities: Modular Decision-Making for LLM Code Review
Rasvik Kudum, Max Corbett, Hitansh Paliwal +3
LLM code reviewers often estimate patch risk and make approval decisions in one prompt. A probability should depend on evidence; costs should determine the action taken from it. We…
CIAware-Bench: Benchmarking Control Intervention Awareness Across Frontier LLMs
Joachim Schaeffer, Thomas Jiralerspong, Alexander Panfilov +4
AI control protocols oversee untrusted models by monitoring their actions and modifying potentially unsafe steps, often using a trusted model. This partially tampers with the untru…
Efficient Causal Graph Discovery Using Large Language Models
Thomas Jiralerspong, Xiaoyin Chen, Yash More +2
We propose a novel framework that leverages LLMs for full causal graph discovery. While previous LLM-based methods have used a pairwise query approach, this requires a quadratic nu…
Noticing the Watcher: LLM Agents Can Infer CoT Monitoring from Blocking Feedback
Thomas Jiralerspong, Flemming Kondrup, Yoshua Bengio
Chain-of-thought (CoT) monitoring is proposed as a method for overseeing the internal reasoning of language-model agents. Prior work has shown that when models are explicitly infor…
Delta-Crosscoder: Robust Crosscoder Model Diffing in Narrow Fine-Tuning Regimes
Aly Kassem, Thomas Jiralerspong, Negar Rostamzadeh +1
Model diffing methods aim to identify how fine-tuning changes a model's internal representations. Crosscoders approach this by learning shared dictionaries of interpretable latent…
Cross-Architecture Model Diffing with Crosscoders: Unsupervised Discovery of Differences Between LLMs
Thomas Jiralerspong, Trenton Bricken
Model diffing, the process of comparing models' internal representations to identify their differences, is a promising approach for uncovering safety-critical behaviors in new mode…