Cover Image

Use Sparse Autoencoders to Discover Unknown Concepts, Not to Act on Known Concepts

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Peng, Kenny, Movva, Rajiv, Kleinberg, Jon, Pierson, Emma, Garg, Nikhil
Format:	Preprint
Published:	2025
Subjects:	Machine Learning Artificial Intelligence Computation and Language Computers and Society
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Sparse Autoencoders for Hypothesis Generation
by: Movva, Rajiv, et al.
Published: (2025)

How Many Features Can a Language Model Store Under the Linear Representation Hypothesis?
by: Garg, Nikhil, et al.
Published: (2026)

Correlated Errors in Large Language Models
by: Kim, Elliot, et al.
Published: (2025)

Topics, Authors, and Institutions in Large Language Model Research: Trends from 17K arXiv Papers
by: Movva, Rajiv, et al.
Published: (2023)

A No Free Lunch Theorem for Human-AI Collaboration
by: Peng, Kenny, et al.
Published: (2024)

Generative AI in Medicine
by: Shanmugam, Divya, et al.
Published: (2024)

Evaluating Adversarial Robustness of Concept Representations in Sparse Autoencoders
by: Li, Aaron J., et al.
Published: (2025)

What's In My Human Feedback? Learning Interpretable Descriptions of Preference Data
by: Movva, Rajiv, et al.
Published: (2025)

Sparse Shift Autoencoders for Identifying Concepts from Large Language Model Activations
by: Joshi, Shruti, et al.
Published: (2025)

Decoding Dark Matter: Specialized Sparse Autoencoders for Interpreting Rare Concepts in Foundation Models
by: Muhamed, Aashiq, et al.
Published: (2024)

Learning Disease Progression Models That Capture Health Disparities
by: Chiang, Erica, et al.
Published: (2024)

From Isolation to Entanglement: When Do Interpretability Methods Identify and Disentangle Known Concepts?
by: Mueller, Aaron, et al.
Published: (2025)

One Example Shown, Many Concepts Known! Counterexample-Driven Conceptual Reasoning in Mathematical LLMs
by: Li, Yinghui, et al.
Published: (2025)

A Bayesian Spatial Model to Correct Under-Reporting in Urban Crowdsourcing
by: Agostini, Gabriel, et al.
Published: (2023)

Position: The Most Expensive Part of an LLM should be its Training Data
by: Kandpal, Nikhil, et al.
Published: (2025)

Values in the Wild: Discovering and Analyzing Values in Real-World Language Model Interactions
by: Huang, Saffron, et al.
Published: (2025)

Cross-Layer Discrete Concept Discovery for Interpreting Language Models
by: Garg, Ankur, et al.
Published: (2025)

COMPL-AI Framework: A Technical Interpretation and LLM Benchmarking Suite for the EU Artificial Intelligence Act
by: Guldimann, Philipp, et al.
Published: (2024)

Emissions and Performance Trade-off Between Small and Large Language Models
by: Garg, Anandita, et al.
Published: (2025)

Language Generation in the Limit
by: Kleinberg, Jon, et al.
Published: (2024)

An Image is Worth Multiple Words: Discovering Object Level Concepts using Multi-Concept Prompt Learning
by: Jin, Chen, et al.
Published: (2023)

Anatomy of a Machine Learning Ecosystem: 2 Million Models on Hugging Face
by: Laufer, Benjamin, et al.
Published: (2025)

Designing Skill-Compatible AI: Methodologies and Frameworks in Chess
by: Hamade, Karim, et al.
Published: (2024)

Concept Tokens: Learning Behavioral Embeddings Through Concept Definitions
by: Sastre, Ignacio, et al.
Published: (2026)

The WMDP Benchmark: Measuring and Reducing Malicious Use With Unlearning
by: Li, Nathaniel, et al.
Published: (2024)

Bayesian Modeling of Zero-Shot Classifications for Urban Flood Detection
by: Franchi, Matt, et al.
Published: (2025)

Policies Permitting LLM Use for Polishing Peer Reviews Are Currently Not Enforceable
by: Saha, Rounak, et al.
Published: (2026)

From Concepts to Components: Concept-Agnostic Attention Module Discovery in Transformers
by: Su, Jingtong, et al.
Published: (2025)

Beyond Single Concept Vector: Modeling Concept Subspace in LLMs with Gaussian Distribution
by: Zhao, Haiyan, et al.
Published: (2024)

Sparse Autoencoder Features for Classifications and Transferability
by: Gallifant, Jack, et al.
Published: (2025)

LangFair: A Python Package for Assessing Bias and Fairness in Large Language Model Use Cases
by: Bouchard, Dylan, et al.
Published: (2025)

Psychological Counseling Ability of Large Language Models
by: Peng, Fangyu, et al.
Published: (2025)

ICE-T: A Multi-Faceted Concept for Teaching Machine Learning
by: Krone, Hendrik, et al.
Published: (2024)

Safe and Certifiable AI Systems: Concepts, Challenges, and Lessons Learned
by: Schweighofer, Kajetan, et al.
Published: (2025)

Incorporating Hierarchical Semantics in Sparse Autoencoder Architectures
by: Muchane, Mark, et al.
Published: (2025)

Exploring Concept Depth: How Large Language Models Acquire Knowledge and Concept at Different Layers?
by: Jin, Mingyu, et al.
Published: (2024)

AlignSAE: Concept-Aligned Sparse Autoencoders
by: Yang, Minglai, et al.
Published: (2025)

PICKT: Practical Interlinked Concept Knowledge Tracing for Personalized Learning using Knowledge Map Concept Relations
by: Lee, Wonbeen, et al.
Published: (2025)

Sparse but Wrong: Incorrect L0 Leads to Incorrect Features in Sparse Autoencoders
by: Chanin, David, et al.
Published: (2025)

Disentangling Heterogeneous Knowledge Concept Embedding for Cognitive Diagnosis on Untested Knowledge
by: Zhang, Miao, et al.
Published: (2024)