mlsanityData to 7 Oct 2026

What you can cram into a single vector: Probing sentence embeddings for linguistic properties

arXiv 1805.01070 ↗First posted 3 May 2018cs.CLcited by 452

Not on the map. Why?

Abstract (arXiv)

Although much effort has recently been devoted to training high-quality sentence embeddings, we still have a poor understanding of what they are capturing. "Downstream" tasks, often based on sentence classification, are commonly used to evaluate the quality of sentence representations. The complexity of the tasks makes it however difficult to infer what kind of information is present in the representations. We introduce here 10 probing tasks designed to capture simple linguistic features of sentences, and we use them to study embeddings generated by three different encoders trained in eight distinct ways, uncovering intriguing properties of both encoders and training methods.

Key facts. Cited by 452 papers in the map (data through 7 Oct 2026). First posted on arXiv 3 May 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 90; 2022: 76; 2023: 79; 2024: 72; 2025: 65; 2026: 70.

452 papers.

Sort:

By area:

  1. 2022Locating and Editing Factual Associations in GPTcited by 2,609
  2. 2022BLOOM: A 176B-Parameter Open-Access Multilingual Language Modelcited by 1,968
  3. 2021Probing Classifiers: Promises, Shortcomings, and Advancescited by 831
  4. 2021Do Vision Transformers See Like Convolutional Neural Networks?cited by 671
  5. 2021Fast Model Editing at Scalecited by 491
  6. 2022Emergent World Representations: Exploring a Sequence Model Trained on a Synthetic Taskcited by 417
  7. 2023Finding Neurons in a Haystack: Case Studies with Sparse Probingcited by 328
  8. 2021Pre-Trained Models: Past, Present and Futurecited by 319
  9. 2023Emergent Linear Representations in World Models of Self-Supervised Sequence Modelscited by 313
  10. 2023Physics of Language Models: Part 3.1, Knowledge Storage and Extractioncited by 264
  11. 2024Assessing the Brittleness of Safety Alignment via Pruning and Low-Rank Modificationscited by 220
  12. 2025AxBench: Steering LLMs? Even Simple Baselines Outperform Sparse Autoencoderscited by 199
38 more
  1. 2023Causal Abstraction: A Theoretical Foundation for Mechanistic Interpretabilitycited by 194
  2. 2023Linearity of Relation Decoding in Transformer Language Modelscited by 190
  3. 2024A Mechanistic Understanding of Alignment Algorithms: A Case Study on DPO and Toxicitycited by 179
  4. 2021Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Littlecited by 176
  5. 2021Post-hoc Interpretability for Neural NLP: A Surveycited by 155
  6. 2023Inspecting and Editing Knowledge Representations in Language Modelscited by 140
  7. 2022Towards Faithful Model Explanation in NLP: A Surveycited by 129
  8. 2022Toward Transparent AI: A Survey on Interpreting the Inner Structures of Deep Neural Networkscited by 126
  9. 2024Black-Box Access is Insufficient for Rigorous AI Auditscited by 123
  10. 2023The Truth is in There: Improving Reasoning in Language Models with Layer-Selective Rank Reductioncited by 120
  11. 2024On Memorization of Large Language Models in Logical Reasoningcited by 114
  12. 2024A Practical Review of Mechanistic Interpretability for Transformer-Based Language Modelscited by 113
  13. 2024Rethinking Interpretability in the Era of Large Language Modelscited by 110
  14. 2023Are Emergent Abilities in Large Language Models just In-Context Learning?cited by 109
  15. 2023Interpretability at Scale: Identifying Causal Mechanisms in Alpacacited by 102
  16. 2023Jump to Conclusions: Short-Cutting Transformers With Linear Transformationscited by 89
  17. 2023Tell Your Model Where to Attend: Post-hoc Attention Steering for LLMscited by 82
  18. 2021Neuron-level Interpretation of Deep NLP Models: A Surveycited by 81
  19. 2024RAVEL: Evaluating Interpretability Methods on Disentangling Language Model Representationscited by 77
  20. 2023Is This the Subspace You Are Looking for? An Interpretability Illusion for Subspace Activation Patchingcited by 75
  21. 2021First Align, then Predict: Understanding the Cross-Lingual Ability of Multilingual BERTcited by 64
  22. 2024How Large Language Models Encode Context Knowledge? A Layer-Wise Probing Studycited by 64
  23. 2021Low-Complexity Probing via Finding Subnetworkscited by 63
  24. 2022Assessing the State of Self-Supervised Human Activity Recognition using Wearablescited by 63
  25. 2023Towards Vision-Language Mechanistic Interpretability: A Causal Tracing Tool for BLIPcited by 62
  26. 2021Codified audio language modeling learns useful representations for music information retrievalcited by 61
  27. 2021On the Pitfalls of Analyzing Individual Neurons in Language Modelscited by 61
  28. 2023Explaining black box text modules in natural language with language modelscited by 61
  29. 2024Steering Knowledge Selection Behaviours in LLMs via SAE-Based Representation Engineeringcited by 60
  30. 2021Counterfactual Interventions Reveal the Causal Effect of Relative Clause Representations on Agreement Predictioncited by 58
  31. 2022minicons: Enabling Flexible Behavioral and Representational Analyses of Transformer Language Modelscited by 58
  32. 2022Probing for the Usage of Grammatical Numbercited by 57
  33. 2024Emergence of a High-Dimensional Abstraction Phase in Language Transformerscited by 56
  34. 2021Provable Limitations of Acquiring Meaning from Ungrounded Form: What Will Future Language Models Understand?cited by 55
  35. 2022CEBaB: Estimating the Causal Effects of Real-World Concepts on NLP Model Behaviorcited by 54
  36. 2023The Curious Case of Hallucinatory (Un)answerability: Finding Truths in the Hidden States of Over-Confident Large Language Modelscited by 54
  37. 2021What all do audio transformer models hear? Probing Acoustic Representations for Language Delivery and its Structurecited by 52
  38. 2022What Do They Capture? -- A Structural Analysis of Pre-Trained Language Models for Source Codecited by 50

50 most cited of 452 All 452 (JSON)