SparseCoding

Navigation page for Library/SparseCoding/ at upstream snapshot 01cda219c1434bb298deda2adcbbdbdfa664e880.

Contents

• Sparse autoencoders find composed features in small toy models
• Block-structured coded retrieval
• Simplex coverage depth and optimality
• Towards Monosemanticity: Decomposing Language Models With Dictionary Learning
• BatchTopK Sparse Autoencoders
• Learning Multi-Level Features with Matryoshka Sparse Autoencoders
• The Convex Geometry of Linear Inverse Problems
• A is for Absorption: Studying Feature Splitting and Absorption in Sparse Autoencoders
• Toy Models of Feature Absorption in SAEs
• How Optimality Structures Sparse Dictionaries: A Theory for Understanding SAE Representations
• Toy Models of Superposition
• Scaling and evaluating sparse autoencoders
• On the Local Correctness of L1 Minimization for Dictionary Learning
• Dictionary Identification – Sparse Matrix-Factorisation via l1-Minimisation
• Sparse Autoencoders Do Not Find Canonical Units of Analysis
• The geometry and identifiability of superposition (MAIS-A3 research draft)
• Toward Identifiable Sparse Autoencoders
• Emergence of simple-cell receptive field properties by learning a sparse code for natural images
• Proximal Algorithms
• A Dominant Diffuse Phase in the Sparse Autoencoder Phase Diagram
• Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders
• Do sparse autoencoders find true features?