mlsanityData to 7 Oct 2026

Disentangling Memory and Reasoning Ability in Large Language Models

arXiv 2411.13504 ↗First posted 20 Nov 2024cs.CLcited by 43

On the map: LLM reasoning → Reasoning benchmarks and evaluation

Abstract (arXiv)

Large Language Models (LLMs) have demonstrated strong performance in handling complex tasks requiring both extensive knowledge and reasoning abilities. However, the existing LLM inference pipeline operates as an opaque process without explicit separation between knowledge retrieval and reasoning steps, making the model's decision-making process unclear and disorganized. This ambiguity can lead to issues such as hallucinations and knowledge forgetting, which significantly impact the reliability of LLMs in high-stakes domains. In this paper, we propose a new inference paradigm that decomposes the complex inference process into two distinct and clear actions: (1) memory recall: which retrieves relevant knowledge, and (2) reasoning: which performs logical steps based on the recalled knowledge. To facilitate this decomposition, we introduce two special tokens memory and reason, guiding the model to distinguish between steps that require knowledge retrieval and those that involve reasoning. Our experiment results show that this decomposition not only improves model performance but also enhances the interpretability of the inference process, enabling users to identify sources of error and refine model responses effectively. The code is available at https://github.com/MingyuJ666/Disentangling-Memory-and-Reasoning.

Key facts. Cited by 43 papers in the map (data through 7 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 38 of its 51 references that are in the map. First posted on arXiv 20 Nov 2024.

Builds on

38 of 51 references.

Closest ancestors · 5

Same direction, newest first

  1. 2024Grokked Transformers are Implicit Reasoners: A Mechanistic Journey to the Edge of Generalization23 May
  2. 2024Understanding and Patching Compositional Reasoning in LLMs22 Feb
  3. 2024Same Task, More Tokens: the Impact of Input Length on the Reasoning Performance of Large Language Models19 Feb
  4. 2024Understanding Reasoning Ability of Language Models From the Perspective of Reasoning Paths Aggregation5 Feb
  5. 2023Enhancing Zero-Shot Chain-of-Thought Reasoning in Large Language Models through Logic23 Sep

Other references · 33

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,084
  3. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  4. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
  5. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,155
  6. 2021TruthfulQA: Measuring How Models Mimic Human Falsehoodscited by 2,919
  7. 2022Emergent Abilities of Large Language Modelscited by 2,483
  8. 2018CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledgecited by 1,793
  9. 2023Graph of Thoughts: Solving Elaborate Problems with Large Language Modelscited by 873
  10. 2020Quantifying Attention Flow in Transformerscited by 759
  11. 2021Did Aristotle Use a Laptop? A Question Answering Benchmark with Implicit Reasoning Strategiescited by 736
  12. 2023Towards Revealing the Mystery behind Chain of Thought: A Theoretical Perspectivecited by 293
21 more
  1. 2023Physics of Language Models: Part 3.1, Knowledge Storage and Extractioncited by 264
  2. 2024To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoningcited by 257
  3. 2023Think before you speak: Training Language Models With Pause Tokenscited by 254
  4. 2023Survey on Factuality in Large Language Models: Knowledge, Retrieval and Domain-Specificitycited by 204
  5. 2023Can LLM-Generated Misinformation Be Detected?cited by 173
  6. 2024Do Large Language Models Latently Perform Multi-Hop Reasoning?cited by 165
  7. 2024The Impact of Reasoning Step Length on Large Language Modelscited by 152
  8. 2024Distilling System 2 into System 1cited by 152
  9. 2024Mind Your Step (by Step): Chain-of-Thought can Reduce Performance on Tasks where Thinking Makes Humans Worsecited by 107
  10. 2023Thread of Thought Unraveling Chaotic Contextscited by 71
  11. 2023Guiding Language Model Reasoning with Planning Tokenscited by 58
  12. 2024Visual Agents as Fast and Slow Thinkerscited by 53
  13. 2018Skeleton-to-Response: Dialogue Generation Guided by Retrieval Memorycited by 49
  14. 2023Deceptive Semantic Shortcuts on Reasoning Chains: How Far Can Models Go without Hallucination?cited by 26
  15. 2024Boosting of Thoughts: Trial-and-Error Problem Solving with Large Language Modelscited by 22
  16. 2024Two-stage Generative Question Answering on Temporal Knowledge Graph Using Large Language Modelscited by 18
  17. 2024M2PT: Multimodal Prompt Tuning for Zero-shot Instruction Learningcited by 16
  18. 2024Towards Reliable Medical Question Answering: Techniques and Challenges in Mitigating Hallucinations in Language Modelscited by 15
  19. 2024Understanding the Interplay between Parametric and Contextual Knowledge for Large Language Modelscited by 14
  20. 2024ToW: Thoughts of Words Improve Reasoning in Large Language Modelscited by 2
  21. 2024Can Language Models Take A Hint? Prompting for Controllable Contextualized Commonsense Inferencecited by 1

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2024: 1; 2025: 25; 2026: 17.

43 papers.

Sort:

By area:

  1. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  2. 2025MemOS: A Memory OS for AI Systemcited by 129
  3. 2025Efficient Reasoning Models: A Surveycited by 93
  4. 2025A Survey on Latent Reasoningcited by 71
  5. 2025Massive Values in Self-Attention Modules are the Key to Contextual Knowledge Understandingcited by 44
  6. 2025Implicit Reasoning in Large Language Models: A Comprehensive Surveycited by 42
  7. 2026Hierarchy-of-Groups Policy Optimization for Long-Horizon Agentic Taskscited by 37
  8. 2025Rethinking Memory in LLM based Agents: Representations, Operations, and Emerging Topicscited by 36
  9. 2025MemSearcher: Training LLMs to Reason, Search and Manage Memory via End-to-End Reinforcement Learningcited by 26
  10. 2025LaDiR: Latent Diffusion Enhances LLMs for Text Reasoningcited by 25
  11. 2025Disentangling Reasoning and Knowledge in Medical Large Language Modelscited by 17
  12. 2025Doxing via the Lens: Revealing Location-related Privacy Leakage on Multi-modal Large Reasoning Modelscited by 14
31 more
  1. 2025General Intelligence Requires Reward-based Pretrainingcited by 12
  2. 2026According to Me: Long-Term Personalized Referential Memory QAcited by 12
  3. 2024Beyond Math: Stories as a Testbed for Memorization-Constrained Reasoning in LLMscited by 11
  4. 2025UORA: Uniform Orthogonal Reinitialization Adaptation in Parameter-Efficient Fine-Tuning of Large Modelscited by 11
  5. 2025Re-Imagining Multimodal Instruction Tuning: A Representation Viewcited by 9
  6. 2025Atomic Skills are the Prerequisite: When Reinforcement Learning Synthesizes Compositional Reasoning, and When It Only Amplifiescited by 8
  7. 2026LaDi-RL: Latent Diffusion Reasoning Prevents Entropy Collapse in Reinforcement Learningcited by 7
  8. 2026LoopMoE: Unifying Iterative Computation with Mixture-of-Experts for Language Modelingcited by 7
  9. 2026SeLaR: Selective Latent Reasoning in Large Language Modelscited by 6
  10. 2025Diagnosing Memorization in Chain-of-Thought Reasoning, One Token at a Timecited by 4
  11. 2025Demystifying Scientific Problem-Solving in LLMs by Probing Knowledge and Reasoningcited by 4
  12. 2025Agent-ScanKit: Unraveling Memory and Reasoning of Multimodal Agents via Sensitivity Perturbationscited by 3
  13. 2026Skill-RAG: Failure-State-Aware Retrieval Augmentation via Hidden-State Probing and Skill Routingcited by 3
  14. 2026Organize then Retrieve: Hierarchical Memory Navigation for Efficient Agentscited by 3
  15. 2025Large Language Models Could Be Rote Learnerscited by 2
  16. 2025Honey, I Shrunk the Language Model: Impact of Knowledge Distillation Methods on Performance and Explainabilitycited by 2
  17. 2025Recall with Reasoning: Chain-of-Thought Distillation for Mamba's Long-Context Memory and Extrapolationcited by 2
  18. 2026Illusions of Confidence? Diagnosing LLM Truthfulness via Neighborhood Consistencycited by 2
  19. 2025COSMIR: Chain Orchestrated Structured Memory for Iterative Reasoning over Long Contextcited by 1
  20. 2025On the Relationship Between the Choice of Representation and In-Context Learningcited by 1
  21. 2026Individual Turing Test: A Case Study of LLM-based Simulation Using Longitudinal Personal Datacited by 1
  22. 2026Specializing Large Models for Oracle Bone Script Interpretation via Component-Grounded Multimodal Knowledge Augmentationcited by 1
  23. 2026Are Rationales Necessary and Sufficient? Tuning LLMs for Explainable Misinformation Detectioncited by 1
  24. 2026The Structural Attention Tax: How Retrieval Format Hijacks In-Context Learning Independent of Contentcited by 1
  25. 2025Chain of Methodologies: Scaling Test Time Computation without Trainingcited by 0
  26. 2025Thinking Before You Speak: A Proactive Test-time Scaling Approachcited by 0
  27. 2026Distilling Reasoning Without Knowledge: A Framework for Reliable LLMscited by 0
  28. 2026IsoSci: A Benchmark of Isomorphic Cross-Domain Science Problems for Evaluating Reasoning versus Knowledge Retrieval in LLMscited by 0
  29. 2026What Makes Agent Memory Useful for Reliable Unanswerable Question Handling?cited by 0
  30. 2026Tracking the Moving Frontier: Long-Short Term Advantage Estimatorcited by 0
  31. 2026X-Tree: Tokenizing Reusable Experience for Efficient Agent Generalizationcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Advancing Reasoning in Large Language Models: Promising Methods and Approaches
  2. 2024On Memorization of Large Language Models in Logical Reasoning
  3. 2024Unveiling Factual Recall Behaviors of Large Language Models through Knowledge Neurons
  4. 2026Enhancing reasoning accuracy in large language models during inference time
  5. 2025Unveiling the Magic of Code Reasoning through Hypothesis Decomposition and Amendment
  6. 2024Beyond Accuracy: Evaluating the Reasoning Behavior of Large Language Models -- A Survey
  7. 2025The Reasoning-Memorization Interplay in Language Models Is Mediated by a Single Direction
  8. 2025Efficient Inference for Large Reasoning Models: A Survey