mlsanityData to 7 Oct 2026

A Survey of Multimodal Mathematical Reasoning: From Perception, Alignment to Reasoning

arXiv 2603.08291 ↗First posted 9 Mar 2026cs.AIcited by 0

On the map: LLM reasoning → Multimodal reasoning

Abstract (arXiv)

Multimodal Mathematical Reasoning (MMR) has recently attracted increasing attention for its capability to solve mathematical problems involving both textual and visual modalities. However, current models still face significant challenges in real-world visual math tasks, often misinterpreting diagrams, failing to align mathematical symbols with visual evidence, or producing inconsistent reasoning steps. Moreover, existing evaluations mainly focus on checking final answers rather than verifying the correctness or executability of each intermediate step. A growing body of recent research addresses these issues by integrating structured perception, explicit alignment, and verifiable reasoning within unified frameworks. To establish a clear roadmap for understanding and comparing different MMR approaches, we systematically review them around four fundamental questions: (1) What to extract from multimodal inputs, (2) How to represent and align textual and visual information, (3) How to perform the reasoning, and (4) How to evaluate the correctness of the overall reasoning process. Finally, we discuss open challenges and share our thoughts on future research directions.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Multimodal reasoning, one of 17 directions in LLM reasoning. Builds on 91 of its 145 references that are in the map. First posted on arXiv 9 Mar 2026.

Builds on

91 of 145 references.

Closest ancestors · 43

Same direction, newest first

  1. 2025VReST: Enhancing Reasoning in Large Vision-Language Models through Tree Search and Self-Reward Mechanism10 Jun
  2. 2025Sherlock: Self-Correcting Reasoning in Vision-Language Models28 May
  3. 2025Integrating Visual Interpretation and Linguistic Reasoning for Math Problem Solving23 May
  4. 2025Think or Not? Selective Reasoning via Reinforcement Learning for Vision-Language Models22 May
  5. 2025VLM-R3: Region Recognition, Reasoning, and Refinement for Enhanced Multimodal Chain-of-Thought22 May
  6. 2025MM-PRM: Enhancing Multimodal Mathematical Reasoning with Scalable Step-Level Supervision19 May
  7. 2025MathCoder-VL: Bridging Vision and Code for Enhanced Multimodal Mathematical Reasoning15 May
  8. 2025Perception, Reason, Think, and Plan: A Survey on Large Multimodal Reasoning Models8 May
  9. 2025VisuoThink: Empowering LVLM Reasoning with Multimodal Tree Search12 Apr
  10. 2025SFT or RL? An Early Investigation into Training R1-Like Reasoning Large Vision-Language Models10 Apr
  11. 2025VL-Rethinker: Incentivizing Self-Reflection of Vision-Language Models with Reinforcement Learning10 Apr
  12. 2025SoTA with Less: MCTS-Guided Sample Selection for Data-Efficient Visual Reasoning Self-Improvement10 Apr
31 more
  1. 2025Rethinking RL Scaling for Vision Language Models: A Transparent, From-Scratch Framework and Comprehensive Evaluation Scheme3 Apr
  2. 2025Math Blind: Failures in Diagram Understanding Undermine Reasoning in MLLMs26 Mar
  3. 2025Mitigating Visual Forgetting via Take-along Visual Conditioning for Multi-modal Long CoT Reasoning17 Mar
  4. 2025R1-Onevision: Advancing Generalized Multimodal Reasoning through Cross-Modal Formalization13 Mar
  5. 2025VisualPRM: An Effective Process Reward Model for Multimodal Reasoning13 Mar
  6. 2025MM-Eureka: Exploring the Frontiers of Multimodal Reasoning with Rule-based Reinforcement Learning10 Mar
  7. 2025Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models9 Mar
  8. 2025The Role of Visual Modality in Multimodal Mathematical Reasoning: Challenges and Insights6 Mar
  9. 2024MAmmoTH-VL: Eliciting Multimodal Reasoning with Instruction Tuning at Scale6 Dec
  10. 2024AtomThink: Multimodal Slow Thinking with Atomic Step Reasoning18 Nov
  11. 2024Thinking Before Looking: Improving Multimodal LLM Reasoning via Mitigating Visual Hallucination15 Nov
  12. 2024LLaVA-CoT: Let Vision Language Models Reason Step-by-Step15 Nov
  13. 2024DynaMath: A Dynamic Visual Benchmark for Evaluating Mathematical Reasoning Robustness of Vision Language Models29 Oct
  14. 2024Polymath: A Challenging Multi-modal Mathematical Reasoning Benchmark6 Oct
  15. 2024CMM-Math: A Chinese Multimodal Math Dataset To Evaluate and Enhance the Mathematics Reasoning of Large Multimodal Models4 Sep
  16. 2024MultiMath: Bridging Visual and Mathematical Reasoning for Large Language Models30 Aug
  17. 2024Math-PUMA: Progressive Upward Multimodal Alignment to Enhance Mathematical Reasoning16 Aug
  18. 2024MathScape: Benchmarking Multimodal Large Language Models in Real-World Mathematical Contexts14 Aug
  19. 2024MAVIS: Mathematical Visual Instruction Tuning with an Automatic Data Engine11 Jul
  20. 2024We-Math: Does Your Large Multimodal Model Achieve Human-like Mathematical Reasoning?1 Jul
  21. 2024Math-LLaVA: Bootstrapping Mathematical Reasoning for Multimodal Large Language Models25 Jun
  22. 2024Describe-then-Reason: Improving Multimodal Mathematical Reasoning through Visual Comprehension Training22 Apr
  23. 2024MM-MATH: Advancing Multimodal Math Evaluation with Process Evaluation and Fine-grained Classification7 Apr
  24. 2024Soft-Prompting with Graph-of-Thought for Multi-modal Representation Learning6 Apr
  25. 2024All in an Aggregated Image for In-Image Learning28 Feb
  26. 2024ChartX & ChartVLM: A Versatile Benchmark and Foundation Model for Complicated Chart Reasoning19 Feb
  27. 2023MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts3 Oct
  28. 2023MM-REACT: Prompting ChatGPT for Multimodal Reasoning and Action20 Mar
  29. 2022DePlot: One-shot visual language reasoning by plot-to-table translation20 Dec
  30. 2022CLEVR-Math: A Dataset for Compositional Language, Visual and Mathematical Reasoning10 Aug
  31. 2022ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning19 Mar

Other references · 48

  1. 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,146
  2. 2024OlympiadBench: A Challenging Benchmark for Promoting AGI with Olympiad-Level Bilingual Multimodal Scientific Problemscited by 1,099
  3. 2017Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learningcited by 679
  4. 2021FinQA: A Dataset of Numerical Reasoning over Financial Datacited by 552
  5. 2015Compositional Semantic Parsing on Semi-Structured Tablescited by 509
  6. 2019TabFact: A Large-scale Dataset for Table-based Fact Verificationcited by 425
  7. 2021TAT-QA: A Question Answering Benchmark on a Hybrid of Tabular and Textual Content in Financecited by 402
  8. 2021Inter-GPS: Interpretable Geometry Problem Solving with Formal Language and Symbolic Reasoningcited by 376
  9. 2022Dynamic Prompt Learning via Policy Gradient for Semi-structured Mathematical Reasoningcited by 347
  10. 2019PlotQA: Reasoning over Scientific Plotscited by 317
  11. 2017FigureQA: An Annotated Figure Dataset for Visual Reasoningcited by 314
  12. 2025R1-VL: Learning to Reason with Multimodal Large Language Models via Step-wise Group Relative Policy Optimizationcited by 296
36 more
  1. 2024Large Language Models for Mathematical Reasoning: Progresses and Challengescited by 283
  2. 2023ToRA: A Tool-Integrated Reasoning Agent for Mathematical Problem Solvingcited by 276
  3. 2021IconQA: A New Benchmark for Abstract Diagram Understanding and Visual Language Reasoningcited by 262
  4. 2024CharXiv: Charting Gaps in Realistic Chart Understanding in Multimodal LLMscited by 229
  5. 2021GeoQA: A Geometric Question Answering Benchmark Towards Multimodal Numerical Reasoningcited by 224
  6. 2023G-LLaVA: Solving Geometric Problem with Multi-Modal Large Language Modelcited by 208
  7. 2023ChartLlama: A Multimodal LLM for Chart Understanding and Generationcited by 158
  8. 2022UniGeo: Unifying Geometry Logical Reasoning via Reformulating Mathematical Expressioncited by 137
  9. 2023Document Understanding Dataset and Evaluation (DUDE)cited by 132
  10. 2025MMVU: Measuring Expert-Level Multi-Discipline Video Understandingcited by 131
  11. 2023GeomVerse: A Systematic Evaluation of Large Models for Geometric Reasoningcited by 107
  12. 2025Why Is Spatial Reasoning Hard for VLMs? An Attention Mechanism Perspective on Focus Areascited by 93
  13. 2023An In-Context Learning Agent for Formal Theorem-Provingcited by 78
  14. 2024MathGenie: Generating Synthetic Data with Question Back-translation for Enhancing Mathematical Reasoning of LLMscited by 72
  15. 2023A Multi-Modal Neural Geometric Solver with Textual Clauses Parsed from Diagramcited by 55
  16. 2024A Survey of Mathematical Reasoning in the Era of Multimodal Large Language Model: Benchmark, Method & Challengescited by 55
  17. 2024CHAMP: A Competition-level Dataset for Fine-Grained Analyses of LLMs' Mathematical Reasoning Capabilitiescited by 43
  18. 2024ErrorRadar: Benchmarking Complex Mathematical Reasoning of Multimodal Large Language Models Via Error Detectioncited by 37
  19. 2024SceMQA: A Scientific College Entrance Level Multimodal Question Answering Benchmarkcited by 35
  20. 2025Fast-Slow Thinking GRPO for Large Vision-Language Model Reasoningcited by 34
  21. 2024GeoGPT4V: Towards Geometric Multi-modal Large Language Models with Geometric Image Generationcited by 31
  22. 2023FormalGeo: An Extensible Formalized Framework for Olympiad Geometric Problem Solvingcited by 26
  23. 2022LogicSolver: Towards Interpretable Math Word Problem Solving with Logical Prompt-enhanced Learningcited by 24
  24. 2024GeoX: Geometric Problem Solving Through Unified Formalized Vision-Language Pre-trainingcited by 23
  25. 2025Solving Inequality Proofs with Large Language Modelscited by 21
  26. 2024Masked Thought: Simply Masking Partial Reasoning Steps Can Improve Mathematical Reasoning Learning of Language Modelscited by 20
  27. 2025TheoremExplainAgent: Towards Video-based Multimodal Explanations for LLM Theorem Understandingcited by 20
  28. 2025Enhancing the Geometric Problem-Solving Ability of Multimodal LLMs via Symbolic-Neural Integrationcited by 18
  29. 2022PGDP5K: A Diagram Parsing Dataset for Plane Geometry Problemscited by 14
  30. 2025TrustGeoGen: Formal-Verified Data Engine for Trustworthy Multi-modal Geometric Problem Solvingcited by 13
  31. 2024VideoCogQA: A Controllable Benchmark for Evaluating Cognitive Abilities in Video-Language Modelscited by 10
  32. 2025Pi-GPS: Enhancing Geometry Problem Solving by Unleashing the Power of Diagrammatic Informationcited by 10
  33. 2025MathFlow: Enhancing the Perceptual Flow of MLLMs for Visual Mathematical Problemscited by 10
  34. 2025ChartQA-X: Generating Explanations for Visual Chart Reasoningcited by 9
  35. 2024Mathador-LM: A Dynamic Benchmark for Mathematical Reasoning on Large Language Modelscited by 8
  36. 2025Generalizable Geometric Image Caption Synthesiscited by 5

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025MV-MATH: Evaluating Multimodal Math Reasoning in Multi-Visual Contexts
  2. 2025VisioMath: Benchmarking Figure-based Mathematical Reasoning in LMMs
  3. 2024Measuring Multimodal Mathematical Reasoning with MATH-Vision Dataset
  4. 2025MathSight: A Benchmark Exploring Have Vision-Language Models Really Seen in University-Level Mathematical Reasoning?
  5. 2026M3-ACE: Rectifying Visual Perception in Multimodal Math Reasoning via Multi-Agentic Context Engineering
  6. 2025What MLLMs Learn about When they Learn about Multimodal Reasoning
  7. 2025Benchmarking Multimodal Mathematical Reasoning with Explicit Visual Dependency
  8. 2025MathCanvas: Intrinsic Visual Chain-of-Thought for Multimodal Mathematical Reasoning