TopoAgent: A Structure-Aware Perception-to-Reasoning Framework for Diagram-to-Graph Topology Extraction with Large Vision-Language Models
On the map: LLM reasoning → Multimodal reasoning
Abstract (arXiv)
Diagram-to-graph topology extraction aims to extract a graph of entities and their connections from a structural diagram. This task remains challenging for current vision-language models because it requires both fine-grained perceptual grounding and topology-aware reasoning with global consistency. We present TopoBench-180, a human-verified benchmark for diagram-to-graph topology extraction, and TopoAgent, a structure-aware perception-to-reasoning framework for reliable topology extraction using large vision-language models. TopoBench-180 contains 180 structural diagrams spanning Web-style and Network-style categories, paired with canonical graph annotations. TopoAgent progressively extracts the target graph by combining grounded perception, global structural priors, canonical node inventory construction, node-centric local-to-global relation reasoning, and topological consistency enforcement. Experiments on TopoBench-180 show that TopoAgent outperforms strong vision-language model baselines and recent visual reasoning frameworks, especially on edge extraction. More broadly, this work fills an important gap in multimodal structured understanding by establishing a benchmark and framework for diagram-to-graph topology extraction. The benchmark and associated resources will be publicly released at https://huggingface.co/datasets/WayneGuo0011/TopoBench-180.
Key facts. Cited by 1 paper in the map (data through 13 Oct 2026). Placed in Multimodal reasoning, one of 17 directions in LLM reasoning. Builds on 19 of its 42 references that are in the map. First posted on arXiv 27 Aug 2026.
Builds on
19 of 42 references.
Closest ancestors · 11
Same direction, newest first
- 2025Small Drafts, Big Verdict: Information-Intensive Visual Reasoning via Speculation23 Oct
- 2025Ovis2.5 Technical Report15 Aug
- 2025Machine Mental Imagery: Empower Multimodal Reasoning with Latent Visual Tokens20 Jun
- 2025Pixel Reasoner: Incentivizing Pixel-Space Reasoning with Curiosity-Driven Reinforcement Learning21 May
- 2025DeepEyes: Incentivizing "Thinking with Images" via Reinforcement Learning20 May
- 2025ChartMuseum: Testing Visual Reasoning Capabilities of Large Vision-Language Models19 May
- 2025KARL: Knowledge-Aware Reasoning and Reinforcement Learning for Knowledge-Intensive Visual Grounding17 Mar
- 2025Multimodal Chain-of-Thought Reasoning: A Comprehensive Survey16 Mar
- 2025Vision-R1: Incentivizing Reasoning Capability in Multimodal Large Language Models9 Mar
- 2024TinyChart: Efficient Chart Understanding with Visual Token Merging and Program-of-Thoughts Learning25 Apr
- 2024Chain-of-Spot: Interactive Reasoning Improves Large Vision-Language Models19 Mar
Other references · 8
- 2025DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learningcited by 6,247
- 2024GPT-4o System Cardcited by 4,536
- 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,191
- 2025InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Modelscited by 1,798
- 2025PaddleOCR 3.0 Technical Reportcited by 138
- 2025MiMo: Unlocking the Reasoning Potential of Language Model -- From Pretraining to Posttrainingcited by 119
- 2025PaddleOCR-VL: Boosting Multilingual Document Parsing via a 0.9B Ultra-Compact Vision-Language Modelcited by 85
- 2025Eagle 2: Building Post-Training Data Strategies from Scratch for Frontier Vision-Language Modelscited by 69
Built on by
Citing papers per year: 2026: 1.
1 paper.
Similar, not cited
Close in wording, no citation either way
- 2025See or Say Graphs: Agent-Driven Scalable Graph Structure Understanding with Vision-Language Models
- 2026Topo-R1: Detecting Topological Anomalies via Vision-Language Models
- 2026ViSR-KGC: Visual Subgraph Reasoning with Vision-Language Models for Multimodal Knowledge Graph Completion
- 2025Graph-to-Vision: Multi-graph Understanding and Reasoning using Vision-Language Models
- 2026GraphVerse: A Comprehensive Visual Graph Reasoning Benchmark for Multimodal Large Language Models
- 2026When Vision Meets Graphs: A Survey on Graph Reasoning and Learning
- 2026ReactBench: A Benchmark for Topological Reasoning in MLLMs on Chemical Reaction Diagrams
- 2024Benchmarking and Improving Large Vision-Language Models for Fundamental Visual Graph Understanding and Reasoning