activity
20242026
collaborators

6 papers

cs.LG2026

Evil Spectra: How Optimisers can Amplify or Suppress Emergent Misalignment

Jason R. Brown, Patrick Leask, Lev McKinney

Emergent misalignment (EM) is a recently discovered phenomenon in LLMs where fine-tuning on a narrow misaligned task, such as writing insecure code, leads to broadly misaligned beh…

cs.AI2026

Think Fast: Estimating No-CoT Task-Completion Time Horizons of Frontier AI Models

Dewi Gould, Francis Rhys Ward, Anders Cairns Woodruff +18

Many efforts to ensure frontier AI models are safe rely on monitoring their chain-of-thought (CoT) reasoning. If models become able to perform sufficiently complex reasoning intern…

cs.CL2025

Minimal and Mechanistic Conditions for Behavioral Self-Awareness in LLMs

Matthew Bozoukov, Matthew Nguyen, Shubkarman Singh +2

Recent studies have revealed that LLMs can exhibit behavioral self-awareness: the ability to accurately describe or predict their own learned behaviors without explicit supervision…

cs.LG2025

Inference-Time Decomposition of Activations (ITDA): A Scalable Approach to Interpreting Large Language Models

Patrick Leask, Neel Nanda, Noura Al Moubayed

Sparse autoencoders (SAEs) are a popular method for decomposing Large Langage Models (LLM) activations into interpretable latents. However, due to their substantial training cost,…

cs.LG2025

Sparse Autoencoders Do Not Find Canonical Units of Analysis

Patrick Leask, Bart Bussmann, Michael Pearce +5

A common goal of mechanistic interpretability is to decompose the activations of neural networks into features: interpretable properties of the input computed by the model. Sparse…

cs.LG2024

BatchTopK Sparse Autoencoders

Bart Bussmann, Patrick Leask, Neel Nanda

Sparse autoencoders (SAEs) have emerged as a powerful tool for interpreting language model activations by decomposing them into sparse, interpretable features. A popular approach i…