AI Safety North East
5 May 2026•Centre for AI Safety, Newcastle University
How reliable are current interpretability methods? Recent work on CoT monitoring and SAEs (co-presented with Toby Pullan)
Research
Published papers I played a key role in creating:
Sparse Autoencoders Can Learn Graded Latents for Relational Composition
Theo Farrell, Patrick Leask, Noura Al Moubayed
Mechanistic Interpretability Workshop at ICML 2026
View on OpenReview
Order by Scale: Relative-Magnitude Relational Composition in Attention-Only Transformers
Theo Farrell, Patrick Leask, Noura Al Moubayed
ResponsibleFM Workshop at NeurIPS 2025
View on OpenReviewReviewing
- ICML 2026 Mechanistic Interpretability Workshop
- NeurIPS 2025 Mechanistic Interpretability Workshop
- NeurIPS 2025 ResponsibleFM Workshop