mlsanityData to 14 Oct 2026

Boosting Large Language Models with Socratic Method for Conversational Mathematics Teaching

arXiv 2407.17349 ↗First posted 24 Jul 2024cs.CLcited by 16

On the map: LLM reasoning → Mathematical reasoning

Abstract (arXiv)

With the introduction of large language models (LLMs), automatic math reasoning has seen tremendous success. However, current methods primarily focus on providing solutions or using techniques like Chain-of-Thought to enhance problem-solving accuracy. In this paper, we focus on improving the capability of mathematics teaching via a Socratic teaching-based LLM (textttSocraticLLM), which guides learners toward profound thinking with clarity and self-discovery via conversation. We collect and release a high-quality mathematical teaching dataset, named textttSocraticMATH, which provides Socratic-style conversations of problems with extra knowledge. Also, we propose a knowledge-enhanced LLM as a strong baseline to generate reliable responses with review, guidance/heuristic, rectification, and summarization. Experimental results show the great advantages of textttSocraticLLM by comparing it with several strong generative models. The codes and datasets are available on urlhttps://github.com/ECNU-ICALK/SocraticMath.

Key facts. Cited by 16 papers in the map (data through 14 Oct 2026). Placed in Mathematical reasoning, one of 17 directions in LLM reasoning. Builds on 26 of its 52 references that are in the map. First posted on arXiv 24 Jul 2024.

Builds on

26 of 52 references.

Closest ancestors · 3

Same direction, newest first

  1. 2023WizardMath: Empowering Mathematical Reasoning for Large Language Models via Reinforced Evol-Instruct18 Aug
  2. 2023MathDial: A Dialogue Tutoring Dataset with Rich Pedagogical Properties Grounded in Math Reasoning Problems23 May
  3. 2022A Survey of Deep Learning for Mathematical Reasoning20 Dec

Other references · 23

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  4. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  5. 2020Scaling Laws for Neural Language Modelscited by 6,586
  6. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  7. 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,182
  8. 2023Qwen Technical Reportcited by 3,344
  9. 2020mT5: A massively multilingual pre-trained text-to-text transformercited by 1,818
  10. 2023WizardLM: Empowering large pre-trained language models to follow complex instructionscited by 1,125
  11. 2023Graph of Thoughts: Solving Elaborate Problems with Large Language Modelscited by 886
  12. 2021Are NLP Models really able to Solve Simple Math Word Problems?cited by 878
11 more
  1. 2021BARTScore: Evaluating Generated Text as Text Generationcited by 722
  2. 2019MathQA: Towards Interpretable Math Word Problem Solving with Operation-Based Formalismscited by 662
  3. 2017Program Induction by Rationale Generation : Learning to Solve and Explain Algebraic Word Problemscited by 650
  4. 2023MAmmoTH: Building Math Generalist Models through Hybrid Instruction Tuningcited by 487
  5. 2023Large Language Models Are State-of-the-Art Evaluators of Translation Qualitycited by 420
  6. 2016Solving General Arithmetic Word Problemscited by 360
  7. 2017Why We Need New Evaluation Metrics for NLGcited by 233
  8. 2023Goat: Fine-tuned LLaMA Outperforms GPT-4 on Arithmetic Taskscited by 84
  9. 2020Ape210K: A Large-Scale and Template-Rich Dataset of Math Word Problemscited by 60
  10. 2022The NCTE Transcripts: A Dataset of Elementary Math Classroom Transcriptscited by 43
  11. 2023The Art of SOCRATIC QUESTIONING: Recursive Thinking with Large Language Modelscited by 22

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 1; 2025: 7; 2026: 8.

16 papers.

Sort:

By area:

  1. 2025One Size doesn't Fit All: A Personalized Conversational Tutoring Agent for Mathematics Instructioncited by 17
  2. 2025A Survey on Mathematical Reasoning and Optimization with Large Language Modelscited by 12
  3. 2025Simultaneous Multi-objective Alignment Across Verifiable and Non-verifiable Rewardscited by 6
  4. 2025Discerning minds or generic tutors? Evaluating instructional guidance capabilities in Socratic LLMscited by 5
  5. 2025Simulating Students with Large Language Models: A Review of Architecture, Mechanisms, and Role Modelling in Education with Generative AIcited by 5
  6. 2024Navigating the Unknown: A Chat-Based Collaborative Interface for Personalized Exploratory Taskscited by 4
  7. 2026EduEVAL-DB: A Role-Based Dataset for Pedagogical Risk Evaluation in Educational Explanationscited by 4
  8. 2026Rewarding How Models Think Pedagogically: Integrating Pedagogical Reasoning and Thinking Rewards for LLMs in Educationcited by 3
  9. 2026AIriskEval-edu: New Dataset for Risk Assessment in AI-mediated K-12 Educational Explanationscited by 3
  10. 2025Teaching LLMs for Step-Level Automatic Math Correction via Reinforcement Learningcited by 2
  11. 2025Evolutionary Reinforcement Learning based AI tutor for Socratic Interdisciplinary Instructioncited by 2
  12. 2026From Solver to Tutor: Evaluating the Pedagogical Intelligence of LLMs with KMP-Benchcited by 2
4 more
  1. 2026BLADE: Better Language Answers through Dialogue and Explanationscited by 0
  2. 2026Generative Artificial Intelligence for Literature Reviewscited by 0
  3. 2026DICE: Decoupling Capability from Intervention Necessity in LLM Tutoringcited by 0
  4. 2026Do LLM Tutors Actually Teach? A Process-Centered Synthesis of the Tutoring Loopcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024LLM Reasoning Engine: Specialized Training for Enhanced Mathematical Reasoning
  2. 2023Large Language Model for Science: A Study on P vs. NP
  3. 2024ChatGLM-Math: Improving Math Problem-Solving in Large Language Models with a Self-Critique Pipeline
  4. 2023Three Questions Concerning the Use of Large Language Models to Facilitate Mathematics Learning
  5. 2022Automatic Generation of Socratic Subquestions for Teaching Math Word Problems
  6. 2024Large Language Models for Mathematical Reasoning: Progresses and Challenges
  7. 2024MathChat: Benchmarking Mathematical Reasoning and Instruction Following in Multi-Turn Interactions
  8. 2024Benchmarking Large Language Models for Math Reasoning Tasks