mlsanityData to 7 Oct 2026

LLM2: Let Large Language Models Harness System 2 Reasoning

arXiv 2412.20372 ↗First posted 29 Dec 2024cs.CL cs.AIcited by 11

On the map: LLM reasoning → Reasoning benchmarks and evaluation

Abstract (arXiv)

Large language models (LLMs) have exhibited impressive capabilities across a myriad of tasks, yet they occasionally yield undesirable outputs. We posit that these limitations are rooted in the foundational autoregressive architecture of LLMs, which inherently lacks mechanisms for differentiating between desirable and undesirable results. Drawing inspiration from the dual-process theory of human cognition, we introduce LLM2, a novel framework that combines an LLM (System 1) with a process-based verifier (System 2). Within LLM2, the LLM is responsible for generating plausible candidates, while the verifier provides timely process-based feedback to distinguish desirable and undesirable outputs. The verifier is trained with a pairwise comparison loss on synthetic process-supervision data generated through our token quality exploration strategy. Empirical results on mathematical reasoning benchmarks substantiate the efficacy of LLM2, exemplified by an accuracy enhancement from 50.3 to 57.8 (+7.5) for Llama3-1B on GSM8K. Furthermore, when combined with self-consistency, LLM2 achieves additional improvements, boosting major@20 accuracy from 56.2 to 70.2 (+14.0).

Key facts. Cited by 11 papers in the map (data through 7 Oct 2026). Placed in Reasoning benchmarks and evaluation, one of 17 directions in LLM reasoning. Builds on 16 of its 37 references that are in the map. First posted on arXiv 29 Dec 2024.

Builds on

16 of 37 references.

References · 16

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2021Training Verifiers to Solve Math Word Problemscited by 9,743
  3. 2021Evaluating Large Language Models Trained on Codecited by 8,053
  4. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,601
  5. 2022Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedbackcited by 3,634
  6. 2023GAIA: a benchmark for General AI Assistantscited by 1,084
  7. 2022Contrastive Decoding: Open-ended Text Generation as Optimizationcited by 598
  8. 2023DoLa: Decoding by Contrasting Layers Improves Factuality in Large Language Modelscited by 367
  9. 2024Distilling System 2 into System 1cited by 152
  10. 2023Rephrase and Respond: Let Large Language Models Ask Better Questions for Themselvescited by 123
  11. 2024A Thorough Examination of Decoding Methods in the Era of LLMscited by 112
  12. 2023System 2 Attention (is something you might need too)cited by 85
4 more
  1. 2024Critical Tokens Matter: Token-Level Contrastive Estimation Enhances LLM's Reasoning Capabilitycited by 80
  2. 2023Contrastive Decoding Improves Reasoning in Large Language Modelscited by 71
  3. 2023Language Models, Agent Models, and World Models: The LAW for Machine Reasoning and Planningcited by 59
  4. 2024Unchosen Experts Can Contribute Too: Unleashing MoE Models' Power by Self-Contrastcited by 19

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 7; 2026: 4.

11 papers.

  1. 2025Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Modelscited by 381
  2. 2025Reasoning RAG via System 1 or System 2: A Survey on Reasoning Agentic Retrieval-Augmented Generation for Industry Challengescited by 19
  3. 2025RL of Thoughts: Navigating LLM Reasoning with Inference-time Reinforcement Learningcited by 16
  4. 2025Reasoning on a Spectrum: Aligning LLMs to System 1 and System 2 Thinkingcited by 14
  5. 2025Enhancing Large Language Model Reasoning with Reward Models: An Analytical Surveycited by 7
  6. 2025A Review of Incorporating Psychological Theories in LLMscited by 5
  7. 2025Adapting Like Humans: A Metacognitive Agent with Test-time Reasoningcited by 4
  8. 2026The role of System 1 and System 2 semantic memory structure in human and LLM biasescited by 1
  9. 2026AMATA: Adaptive Multi-Agent Trajectory Alignment for Knowledge-Intensive Question Answeringcited by 1
  10. 2026From Abstract to Contextual: What LLMs Still Cannot Do in Mathematicscited by 0
  11. 2026Standing on the Shoulders of Giants: Stabilized Knowledge Distillation for Cross--Language Code Clone Detectioncited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025From System 1 to System 2: A Survey of Reasoning Large Language Models
  2. 2025Advancing Reasoning in Large Language Models: Promising Methods and Approaches
  3. 2024Beyond Accuracy: Evaluating the Reasoning Behavior of Large Language Models -- A Survey
  4. 2023Chain-of-Thought Hub: A Continuous Effort to Measure Large Language Models' Reasoning Performance
  5. 2025A Survey of Scaling in Large Language Model Reasoning
  6. 2025Reasoning on a Budget: A Survey of Adaptive and Controllable Test-Time Compute in LLMs
  7. 2026Opening the Black Box: A Survey on the Mechanisms of Multi-Step Reasoning in Large Language Models
  8. 2024Assessing the Emergent Symbolic Reasoning Abilities of Llama Large Language Models