Compact Proofs of Model Performance via Mechanistic Interpretability

We propose using mechanistic interpretability -- techniques for reverse engineering model weights into human-interpretable algorithms -- to derive and compactly prove formal guarantees on model performance. We prototype this approach by formally proving accuracy lower bounds for a small transformer...

Ausführliche Beschreibung

Bibliographische Detailangaben

Hauptverfasser:	Gross, Jason, Agrawal, Rajashree, Kwa, Thomas, Ong, Euan, Yip, Chun Hei, Gibson, Alex, Noubir, Soufiane, Chan, Lawrence
Format:	Artikel
Sprache:	eng
Schlagworte:	Computer Science - Learning Computer Science - Logic in Computer Science
Online-Zugang:	Volltext bestellen
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!