model representations
- Brain-Informed Fine-Tuning for Improved Multilingual Understanding in Language Models
- Learning to Learn with Contrastive Meta-Objective
- Learning to Learn with Contrastive Meta-Objective
- Overcoming Sparsity Artifacts in Crosscoders to Interpret Chat-Tuning
- Projecting Assumptions: The Duality Between Sparse Autoencoders and Concept Geometry