Modelpedia
Work in progress
About
Findings
Models
Concepts
Methods
Datasets
Sources
Light
Dark
HyperDAS: Towards Automating Mechanistic Interpretability with Hypernetworks
2025-01-22
· ICLR 2025 Poster ·
anchor
Findings
IC-003
HyperDAS dynamically selects intervention tokens and learns linear subspaces in Llama3-8b that mediate entity attributes.