No article text was recorded for this version.
sparse feature decomposition
Version 1 of 11,026 papers1,405 conceptsfrozen 2026-10-04
Papers in this version
- Blog: Interpretability as Compression: Reconsidering SAE Explanations of Neural Activations
- Blog: Breaking Bad Tokens: Detoxification of LLMs Using Sparse Autoencoders
- Blog: Model Directions, Not Words: Mechanistic Topic Models Using Sparse Autoencoders
- Blog: XNNTab - Interpretable Neural Networks for Tabular Data using Sparse Autoencoders
- Blog: Beyond Label Attention: Transparency in Language Models for Automated Medical Coding via Dictionary Learning
- Blog: Transcoders Find Interpretable LLM Feature Circuits
- Blog: Sparse Autoencoders are Topic Models
- Blog: Towards Interpretable Deep Neural Networks for Tabular Data
- Blog: CRISP: Persistent Concept Unlearning via Sparse Autoencoders
- Blog: SAEs Can Improve Unlearning: Dynamic Sparse Autoencoder Guardrails for Precision Unlearning in LLMs
- Blog: Mechanistic Interpretability of Antibody Language Models Using SAEs
- Blog: A Sparsity-promoting Dictionary Model for Variational Autoencoders
- Blog: Steering Large Language Model Activations in Sparse Spaces
- Blog: Disentangling Polysemantic Channels in Convolutional Neural Networks
- Blog: Denoising Concept Vectors with Sparse Autoencoders for Improved Language Model Steering
- Blog: A Compression Objective and a Cycle Loss for Neural Image Compression
- Blog: Interpreting and Steering Protein Language Models through Sparse Autoencoders
- Blog: Identifiable Variational Autoencoders via Sparse Decoding
- Blog: The Knowledge Microscope: Features as Better Analytical Lenses than Neurons
- Blog: Unpacking SDXL Turbo: Interpreting Text-to-Image Models with Sparse Autoencoders
- Blog: Hebbian Learning through the Lens of Sparse Autoencoders
- Blog: Discovering Interpretable Biological Concepts in Single-cell RNA-seq Foundation Models
- Blog: Sparse Autoencoders Reveal Interpretable Structure in Small Gene Language Models
- Blog: Large Multi-modal Models Can Interpret Features in Large Multi-modal Models
- Blog: Sparse Autoencoders for Low-$N$ Protein Function Prediction and Design
- Blog: Scaling sparse feature circuit finding for in-context learning
- Blog: A Fault Diagnosis Method for Rolling Bearings Based on Parameter Transfer Learning under Imbalance Data Sets
- Blog: Feature-Level Insights into Artificial Text Detection with Sparse Autoencoders
- Blog: Label-Free Mitigation of Spurious Correlations in VLMs using Sparse Autoencoders
- Blog: Model Unlearning via Sparse Autoencoder Subspace Guided Projections
- Blog: Interpret and Control Dense Retrieval with Sparse Latent Features
- Blog: Sparse Autoencoders Do Not Find Canonical Units of Analysis
- Blog: Use Sparse Autoencoders to Discover Unknown Concepts, Not to Act on Known Concepts
- Blog: From Black Box to Biomarker: Sparse Autoencoders for Interpreting Speech Models of Parkinson's Disease
- Blog: Precise and Interpretable Editing of Code Knowledge in Large Language Models
- Blog: Simple Sparsification Improves Sparse Denoising Autoencoders in Denoising Highly Corrupted Images
- Blog: Sparse Autoencoders Do Not Find Canonical Units of Analysis
- Blog: Sparse Autoencoders Learn Monosemantic Features in Vision-Language Models
- Blog: Invariance-adapted decomposition and Lasso-type contrastive learning
- Blog: From superposition to sparse codes: interpretable representations in neural networks
- Blog: Transformer visualization via dictionary learning: contextualized embedding as a linear superposition of transformer factors
- Blog: Precise In-Parameter Concept Erasure in Large Language Models
- Blog: From Mechanistic Interpretability to Mechanistic Biology: Training, Evaluating, and Interpreting Sparse Autoencoders on Protein Language Models
- Blog: Interpretable VAEs for nonlinear group factor analysis
- Blog: Safe-SAIL: Towards a Fine-grained Safety Landscape of Large Language Models via Sparse Autoencoder Interpretation Framework
- Blog: SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability
- Blog: SAFE: A Sparse Autoencoder-Based Framework for Robust Query Enrichment and Hallucination Mitigation in LLMs
- Blog: Probing by Analogy: Decomposing Probes into Activations for Better Interpretability and Inter-Model Generalization
- Blog: Data Whitening Improves Sparse Autoencoder Learning
- Blog: ConceptScope: Characterizing Dataset Bias via Disentangled Visual Concepts
- Blog: Steering Large Language Model Activations in Sparse Spaces
- Blog: Sparse Autoencoders Find Highly Interpretable Features in Language Models
- Blog: SAE-FiRE: Enhancing Earnings Surprise Predictions Through Sparse Autoencoder Feature Selection
- Blog: Towards Interpretable Scientific Foundation Models: Sparse Autoencoders for Disentangling Dense Embeddings of Scientific Concepts
- Blog: Interpreting large text-to-image diffusion models with dictionary learning
- Blog: Sparse Fine-Tuning of Transformers for Generative Tasks
- Blog: Interpretable Company Similarity with Sparse Autoencoders
- Blog: Interpreting the linear structure of vision-language model embedding spaces
- Blog: CytoSAE: Interpretable Cell Embeddings for Hematology
- Blog: Sparse Autoencoder as a Zero-Shot Classifier for Concept Erasing in Text-to-Image Diffusion Models
Showing 60 of 1,026 papers.
History & citation
Current · Permanent link
- Disentangling Model Activations with Sparse Autoencoders merged into this direction (77% overlap)
- Automated Semantic Labeling of Latent Features merged into this direction (63% overlap)
- Resolving Neural Polysemanticity via Sparse Autoencoders merged into this direction (69% overlap)
- Causal Feature Steering with Sparse Autoencoders merged into this direction (73% overlap)
- Hierarchical Feature Absorption in Sparse Autoencoders merged into this direction (50% overlap)
- Scaling Feature Interpretability with Automated Explanation Pipelines merged into this direction (57% overlap)
- Decomposing Neural Representations with Sparse Autoencoders merged into this direction (95% overlap)
- Sparse Dictionary Learning for Exhaustive Feature Discovery merged into this direction (67% overlap)
- Sparse Constraints for Identifiable and Interpretable Latent Spaces merged into this direction (52% overlap)
- Monosemantic Feature Discovery in Polysemantic Neural Activations merged into this direction (57% overlap)
Cite this direction
Latest version (a living page):
@misc{lacuna_sparse_feature_decomposition_ea1d3395,
title = {sparse feature decomposition},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/sparse-feature-decomposition/rdr_ea1d33954b8f4de7845eaf52278fdc80},
note = {Latest version of a living research direction; cite a version for a fixed text. Accessed 2026-11-21},
}This exact version (version 1, frozen text):
@misc{lacuna_sparse_feature_decomposition_ea1d3395_v1,
title = {sparse feature decomposition},
author = {{Lacuna}},
howpublished = {Lacuna research direction},
year = {2026},
url = {https://lacuna.tiptreesystems.com/direction/sparse-feature-decomposition/rdr_ea1d33954b8f4de7845eaf52278fdc80/v/1},
note = {Version 1, frozen 2026-10-04. Accessed 2026-11-21},
}