Mechanistic Interpretability

Interpretabilityactive

Reverse-engineering neural networks to identify circuits, features, and algorithms that explain behavior.

Organizations

Key Papers

First Proposed: 2020 (Olah et al., Anthropic)

Cluster: Interpretability

Organizations2

Organization	Role
Anthropic	pioneer
Google DeepMind	active

SEMINAL

Olah et al.2020

SEMINAL

Anthropic2024

Name	Status	Papers
Finding Feature RepresentationsResearch beyond SAEs into alternative methods for identifying latent features in model activations.	emerging	0
Sparse AutoencodersUsing sparse dictionary learning to extract interpretable features from model activations at scale.	active	1
Toy Models for InterpretabilitySmall simplified model proxies that capture key deep learning dynamics for interpretability research.	active	0