mlsanityData to 7 Oct 2026

Attention Lens: A Tool for Mechanistically Interpreting the Attention Head Information Retrieval Mechanism

arXiv 2310.16270 ↗First posted 25 Oct 2023cs.CL cs.AI cs.LGcited by 21

Not on the map. Why?

Abstract (arXiv)

Transformer-based Large Language Models (LLMs) are the state-of-the-art for natural language tasks. Recent work has attempted to decode, by reverse engineering the role of linear layers, the internal mechanisms by which LLMs arrive at their final predictions for text completion tasks. Yet little is known about the specific role of attention heads in producing the final token prediction. We propose Attention Lens, a tool that enables researchers to translate the outputs of attention heads into vocabulary tokens via learned attention-head-specific transformations called lenses. Preliminary findings from our trained lenses indicate that attention heads play highly specialized roles in language models. The code for Attention Lens is available at github.com/msakarvadia/AttentionLens.

Key facts. Cited by 21 papers in the map (data through 7 Oct 2026). Builds on 13 of its 18 references that are in the map. First posted on arXiv 25 Oct 2023.

Builds on

13 of 18 references.

References · 13

  1. 2022Locating and Editing Factual Associations in GPTcited by 2,609
  2. 2020Transformer Feed-Forward Layers Are Key-Value Memoriescited by 1,307
  3. 2022Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 smallcited by 1,130
  4. 2015Aligning Books and Movies: Towards Story-like Visual Explanations by Watching Movies and Reading Bookscited by 1,041
  5. 2023Steering Language Models With Activation Engineeringcited by 824
  6. 2022Ignore Previous Prompt: Attack Techniques For Language Modelscited by 760
  7. 2023Eliciting Latent Predictions from Transformers with the Tuned Lenscited by 579
  8. 2023Dissecting Recall of Factual Associations in Auto-Regressive Language Modelscited by 512
  9. 2023PMET: Precise Model Editing in a Transformercited by 170
  10. 2023The Hydra Effect: Emergent Self-repair in Language Model Computationscited by 145
  11. 2022Analyzing Transformers in Embedding Spacecited by 131
  12. 2023Backdoor Attacks for In-Context Learning with Language Modelscited by 92
1 more
  1. 2023Memory Injections: Correcting Multi-Hop Reasoning Failures during Inference in Transformer-Based Language Modelscited by 21

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 9; 2025: 6; 2026: 6.

21 papers.

Sort:

By area:

  1. 2024Mechanistic Interpretability for AI Safety -- A Reviewcited by 415
  2. 2024A Practical Review of Mechanistic Interpretability for Transformer-Based Language Modelscited by 113
  3. 2024Cutting Off the Head Ends the Conflict: A Mechanism for Interpreting and Mitigating Knowledge Conflicts in Language Modelscited by 87
  4. 2024A Primer on the Inner Workings of Transformer-based Language Modelscited by 83
  5. 2024Finding Transformer Circuits with Edge Pruningcited by 58
  6. 2024A Concept-Based Explainability Framework for Large Multimodal Modelscited by 39
  7. 2024On Behalf of the Stakeholders: Trends in NLP Model Interpretability in the Era of LLMscited by 24
  8. 2026Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Modelscited by 22
  9. 2025Head Pursuit: Probing Attention Specialization in Multimodal Transformerscited by 16
  10. 2025Analyzing Finetuning Representation Shift for Multimodal LLMs Steeringcited by 15
  11. 2025Decoupling Positional and Symbolic Attention Behavior in Transformerscited by 6
  12. 2024Trillion Parameter AI Serving Infrastructure for Scientific Discovery: A Survey and Visioncited by 5
9 more
  1. 2026Counting Circuits: Mechanistic Interpretability of Visual Reasoning in Large Vision-Language Modelscited by 3
  2. 2026Legible-by-Construction: Attention and End-to-End Transformerscited by 2
  3. 2024Universal Response and Emergence of Induction in LLMscited by 1
  4. 2025Hyperdimensional Probe: Decoding LLM Representations via Vector Symbolic Architecturescited by 1
  5. 2025InTraVisTo: Inside Transformer Visualisation Toolcited by 0
  6. 2025Interpret, prune and distill Donut : towards lightweight VLMs for VQA on documentcited by 0
  7. 2026KA2L: A Knowledge-Aware Active Learning Framework for LLMscited by 0
  8. 2026Training, Reading, and Editing Legible Transformerscited by 0
  9. 2026Interpreting Language Model Hidden States at Scalecited by 0