mlsanityData to 7 Oct 2026

Linear Algebraic Structure of Word Senses, with Applications to Polysemy

arXiv 1601.03764 ↗First posted 14 Jan 2016cs.CL cs.LG stat.MLcited by 165

Not on the map. Why?

Abstract (arXiv)

Word embeddings are ubiquitous in NLP and information retrieval, but it is unclear what they represent when the word is polysemous. Here it is shown that multiple word senses reside in linear superposition within the word embedding and simple sparse coding can recover vectors that approximately capture the senses. The success of our approach, which applies to several embedding methods, is mathematically explained using a variant of the random walk on discourses model (Arora et al., 2016). A novel aspect of our technique is that each extracted word sense is accompanied by one of about 2000 "discourse atoms" that gives a succinct description of which other words co-occur with that word sense. Discourse atoms can be of independent interest, and make the method potentially more useful. Empirical tests are used to verify and support the theory.

Key facts. Cited by 165 papers in the map (data through 7 Oct 2026). First posted on arXiv 14 Jan 2016.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 17; 2022: 6; 2023: 18; 2024: 33; 2025: 54; 2026: 37.

165 papers.

Sort:

By area:

  1. 2024Mechanistic Interpretability for AI Safety -- A Reviewcited by 415
  2. 2023Finding Neurons in a Haystack: Case Studies with Sparse Probingcited by 328
  3. 2023Emergent Linear Representations in World Models of Self-Supervised Sequence Modelscited by 313
  4. 2024The Geometry of Categorical and Hierarchical Concepts in Large Language Modelscited by 151
  5. 2024Llama Scope: Extracting Millions of Features from Llama-3.1-8B with Sparse Autoencoderscited by 133
  6. 2021Toward Understanding the Feature Learning Process of Self-supervised Contrastive Learningcited by 119
  7. 2024Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE)cited by 116
  8. 2022Modality Competition: What Makes Joint Training of Multi-modal Network Fail in Deep Learning? (Provably)cited by 115
  9. 2025Are Sparse Autoencoders Useful? A Case Study in Sparse Probingcited by 109
  10. 2021Transformer visualization via dictionary learning: contextualized embedding as a linear superposition of transformer factorscited by 106
  11. 2024Automatically Interpreting Millions of Features in Large Language Modelscited by 106
  12. 2025LLM Social Simulations Are a Promising Research Methodcited by 97
38 more
  1. 2024Improving Steering Vectors by Targeting Sparse Autoencoder Featurescited by 86
  2. 2024Towards Principled Evaluations of Sparse Autoencoders for Interpretability and Controlcited by 83
  3. 2024RAVEL: Evaluating Interpretability Methods on Disentangling Language Model Representationscited by 77
  4. 2024On the Origins of Linear Representations in Large Language Modelscited by 76
  5. 2025Persona Features Control Emergent Misalignmentcited by 76
  6. 2023Is This the Subspace You Are Looking for? An Interpretability Illusion for Subspace Activation Patchingcited by 75
  7. 2025A Survey on Sparse Autoencoders: Interpreting the Internal Mechanisms of Large Language Modelscited by 73
  8. 2025Sparse Autoencoders Learn Monosemantic Features in Vision-Language Modelscited by 70
  9. 2024Identifying Functionally Important Features with End-to-End Sparse Dictionary Learningcited by 64
  10. 2024InterPLM: Discovering Interpretable Features in Protein Language Models via Sparse Autoencoderscited by 64
  11. 2025Projecting Assumptions: The Duality Between Sparse Autoencoders and Concept Geometrycited by 58
  12. 2023From Language Modeling to Instruction Following: Understanding the Behavior Shift in LLMs after Instruction Tuningcited by 56
  13. 2024Measuring Progress in Dictionary Learning for Language Model Interpretability with Board Game Modelscited by 51
  14. 2025The Origins of Representation Manifolds in Large Language Modelscited by 46
  15. 2023Codebook Features: Sparse and Discrete Interpretability for Neural Networkscited by 45
  16. 2024Usable XAI: 10 Strategies Towards Exploiting Explainability in the LLM Eracited by 45
  17. 2025From Flat to Hierarchical: Extracting Sparse Representations with Matching Pursuitcited by 45
  18. 2023Exploring Memorization in Fine-tuned Language Modelscited by 44
  19. 2024The Missing Curve Detectors of InceptionV1: Applying Sparse Autoencoders to InceptionV1 Early Visioncited by 37
  20. 2024Reconciling Kaplan and Chinchilla Scaling Lawscited by 36
  21. 2024Mathematical Models of Computation in Superpositioncited by 34
  22. 2025The Non-Linear Representation Dilemma: Is Causal Abstraction Enough for Mechanistic Interpretability?cited by 34
  23. 2023Reverse Engineering Self-Supervised Learningcited by 33
  24. 2023Which Features are Learnt by Contrastive Learning? On the Role of Simplicity Bias in Class Collapse and Feature Suppressioncited by 30
  25. 2025Superposition Yields Robust Neural Scalingcited by 30
  26. 2024Out-of-distribution generalization via composition: a lens through induction heads in Transformerscited by 28
  27. 2024Dictionary Learning Improves Patch-Free Circuit Discovery in Mechanistic Interpretability: A Case Study on Othello-GPTcited by 27
  28. 2025Taming Knowledge Conflicts in Language Modelscited by 27
  29. 2026When Does LeJEPA Learn a World Model?cited by 27
  30. 2026Do Sparse Autoencoders Capture Concept Manifolds?cited by 25
  31. 2024The Quest for the Right Mediator: Surveying Mechanistic Interpretability Through the Lens of Causal Mediation Analysiscited by 24
  32. 2026Manifold Steering Reveals the Shared Geometry of Neural Network Representation and Behaviorcited by 22
  33. 2024On the Complexity of Neural Computation in Superpositioncited by 21
  34. 2024Mechanistic Permutability: Match Features Across Layerscited by 21
  35. 2025Interpreting and Steering LLMs with Mutual Information-based Explanations on Sparse Autoencoderscited by 21
  36. 2025The Geometry of Reasoning: Flowing Logics in Representation Spacecited by 21
  37. 2023Discovering Universal Geometry in Embeddings with ICAcited by 19
  38. 2024PaCE: Parsimonious Concept Engineering for Large Language Modelscited by 19

50 most cited of 165 All 165 (JSON)