Skip to content
NeurIPS 2026 planner

Poster · Sydney

LoRAcles: Self-Supervised Weight-Space Interpretability at Scale

When
Wed 9 Dec 2026, 10:00–13:00 AEDT
Where
Hall 1-4, ICC Sydney
Exhibition Centre, Level 1
Session
Sydney Poster Session 3
Oral
Oral of this paper
Authors
Celeste De Schamphelaere, Jan Bauer, Neel Nanda, Euan Ong
Organisations
MATS, HebrewU, University College London, Google DeepMind, Anthropic
QRS classification
Language models, Safety, fairness and privacy, Representation and transfer learning

Open in scheduleOfficial page