mlsanityData to 14 Oct 2026

System 2 Attention (is something you might need too)

arXiv 2311.11829 ↗First posted 20 Nov 2023cs.CL cs.AI cs.LGcited by 89

Not on the map. Why?

Cited by 37 in LLM reasoning, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

Soft attention in Transformer-based Large Language Models (LLMs) is susceptible to incorporating irrelevant information from the context into its latent representations, which adversely affects next token generations. To help rectify these issues, we introduce System 2 Attention (S2A), which leverages the ability of LLMs to reason in natural language and follow instructions in order to decide what to attend to. S2A regenerates the input context to only include the relevant portions, before attending to the regenerated context to elicit the final response. In experiments, S2A outperforms standard attention-based LLMs on three tasks containing opinion or irrelevant information, QA, math word problems and longform generation, where S2A increases factuality and objectivity, and decreases sycophancy.

Key facts. Cited by 89 papers in the map (data through 14 Oct 2026). Builds on 26 of its 30 references that are in the map. First posted on arXiv 20 Nov 2023.

Builds on

26 of 30 references.

References · 26

  1. 2017Attention Is All You Needcited by 60,361
  2. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  3. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  4. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,548
  5. 2022Large Language Models are Zero-Shot Reasonerscited by 4,374
  6. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
  7. 2022Constitutional AI: Harmlessness from AI Feedbackcited by 3,050
  8. 2019The Curious Case of Neural Text Degenerationcited by 2,916
  9. 2015Show, Attend and Tell: Neural Image Caption Generation with Visual Attentioncited by 1,423
  10. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 1,407
  11. 2022Measuring and Narrowing the Compositionality Gap in Language Modelscited by 1,075
  12. 2023Towards Understanding Sycophancy in Language Modelscited by 1,040
14 more
  1. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 841
  2. 2022Discovering Language Model Behaviors with Model-Written Evaluationscited by 836
  3. 2020Recipes for building an open-domain chatbotcited by 593
  4. 2017Adversarial Examples for Evaluating Reading Comprehension Systemscited by 571
  5. 2019Neural Text Generation with Unlikelihood Trainingcited by 503
  6. 2014Memory Networkscited by 363
  7. 2014Recurrent Models of Visual Attentioncited by 349
  8. 2023Chain-of-Verification Reduces Hallucination in Large Language Modelscited by 324
  9. 2023Simple synthetic data reduces sycophancy in large language modelscited by 187
  10. 2021Masked Language Modeling and the Distributional Hypothesis: Order Word Matters Pre-training for Littlecited by 177
  11. 2023Rephrase and Respond: Let Large Language Models Ask Better Questions for Themselvescited by 123
  12. 2020UnNatural Language Inferencecited by 69
  13. 2023Context Aware Query Rewriting for Text Rankers using LLMcited by 14
  14. 2023Improving Zero-shot Reader by Reducing Distractions from Irrelevant Documents in Open-Domain Question Answeringcited by 2

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 44; 2025: 25; 2026: 20.

89 papers.

Sort:

By area:

  1. 2024A Systematic Survey of Prompt Engineering in Large Language Models: Techniques and Applicationscited by 431
  2. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 384
  3. 2024Can LLMs Generate Novel Research Ideas? A Large-Scale Human Study with 100+ NLP Researcherscited by 363
  4. 2024RAFT: Adapting Language Model to Domain Specific RAGcited by 262
  5. 2025LightMem: Lightweight and Efficient Memory-Augmented Generationcited by 180
  6. 2024Distilling System 2 into System 1cited by 159
  7. 2024A Human-Inspired Reading Agent with Gist Memory of Very Long Contextscited by 146
  8. 2025Monte Carlo Tree Search for Comprehensive Exploration in LLM-Based Automatic Heuristic Designcited by 109
  9. 2024Sycophancy in Large Language Models: Causes and Mitigationscited by 101
  10. 2025WoW: Towards a World omniscient World model Through Embodied Interactioncited by 67
  11. 2024Demystifying Chains, Trees, and Graphs of Thoughtscited by 63
  12. 2024DocFinQA: A Long-Context Financial Reasoning Datasetcited by 60
38 more
  1. 2024Measuring and Controlling Instruction (In)Stability in Language Model Dialogscited by 53
  2. 2025Does Thinking More always Help? Mirage of Test-Time Scaling in Reasoning Modelscited by 53
  3. 2024A Survey of Prompt Engineering Methods in Large Language Models for Different NLP Taskscited by 47
  4. 2024Thinking LLMs: General Instruction Following with Thought Generationcited by 45
  5. 2024Dualformer: Controllable Fast and Slow Thinking by Learning with Randomized Reasoning Tracescited by 43
  6. 2025A Survey of Test-Time Compute: From Intuitive Inference to Deliberate Reasoningcited by 40
  7. 2024Bias-Augmented Consistency Training Reduces Biased Reasoning in Chain-of-Thoughtcited by 39
  8. 2024LLM Self-Correction with DeCRIM: Decompose, Critique, and Refine for Enhanced Following of Instructions with Multiple Constraintscited by 35
  9. 2024CoGenesis: A Framework Collaborating Large and Small Language Models for Secure Context-Aware Instruction Followingcited by 33
  10. 2024Position: Theory of Mind Benchmarks are Broken for Large Language Modelscited by 31
  11. 2025Leaky Thoughts: Large Reasoning Models Are Not Private Thinkerscited by 28
  12. 2024ALR2: A Retrieve-then-Reason Framework for Long-context Question Answeringcited by 26
  13. 2024LG-Traj: LLM Guided Pedestrian Trajectory Predictioncited by 25
  14. 2024Understanding Long Videos with Multimodal Language Modelscited by 25
  15. 2024To Trust or Not to Trust? Enhancing Large Language Models' Situated Faithfulness to External Contextscited by 25
  16. 2024Perceptions to Beliefs: Exploring Precursory Inferences for Theory of Mind in Large Language Modelscited by 22
  17. 2024LLM Task Interference: An Initial Study on the Impact of Task-Switch in Conversational Historycited by 19
  18. 2024Conceptual and Unbiased Reasoning in Language Modelscited by 19
  19. 2024Chain of Logic: Rule-Based Reasoning with Large Language Modelscited by 18
  20. 2024Verbalized Machine Learning: Revisiting Machine Learning with Language Modelscited by 17
  21. 2025Reasoning on a Spectrum: Aligning LLMs to System 1 and System 2 Thinkingcited by 16
  22. 2024Social Skill Training with Large Language Modelscited by 15
  23. 2026Jev-Mem: System-One-Controlled Agentic Memory for Efficient AI Agentscited by 14
  24. 2024LLM2: Let Large Language Models Harness System 2 Reasoningcited by 12
  25. 2024Exploring and Controlling Diversity in LLM-Agent Conversationcited by 12
  26. 2025Generating Symbolic World Models via Test-time Scaling of Large Language Modelscited by 12
  27. 2024Large Language Models for Social Networks: Applications, Challenges, and Solutionscited by 11
  28. 2024Refined Direct Preference Optimization with Synthetic Data for Behavioral Alignment of LLMscited by 10
  29. 2024Can't Remember Details in Long Documents? You Need Some R&Rcited by 10
  30. 2025The First Impression Problem: Internal Bias Triggers Overthinking in Reasoning Modelscited by 10
  31. 2026InfMem: Learning System-2 Memory Control for Long-Context Agentcited by 9
  32. 2024LOGIN: A Large Language Model Consulted Graph Neural Network Training Frameworkcited by 8
  33. 2024Are LLMs Rational Investors? A Study on Detecting and Reducing the Financial Bias in LLMscited by 7
  34. 2024System-2 Mathematical Reasoning via Enriched Instruction Tuningcited by 7
  35. 2025Unveiling the Magic of Code Reasoning through Hypothesis Decomposition and Amendmentcited by 6
  36. 2024MultiMax: Sparse and Multi-Modal Attention Learningcited by 5
  37. 2024Pixels and Predictions: Potential of GPT-4V in Meteorological Imagery Analysis and Forecast Communicationcited by 4
  38. 2024CItruS: Chunked Instruction-aware State Eviction for Long Sequence Modelingcited by 4

50 most cited of 89 All 89 (JSON)