# SciForma: Structure-Faithful Generation of Scientific Diagrams

arXiv: https://arxiv.org/abs/2607.18091 · HTML: https://mlsanity.com/paper/2607.18091/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 32 of its 66 references that are in the map.
- First posted on arXiv 20 Jul 2026.

## Builds on (most cited first, 30 of 32)

- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Training Diffusion Models with Reinforcement Learning](https://mlsanity.com/paper/2305.13301/) (2023)
- [Flow-GRPO: Training Flow Matching Models via Online RL](https://mlsanity.com/paper/2505.05470/) (2025)
- [DanceGRPO: Unleashing GRPO on Visual Generation](https://mlsanity.com/paper/2505.07818/) (2025)
- [Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer](https://mlsanity.com/paper/2511.22699/) (2025)
- [DiffusionNFT: Online Diffusion Reinforcement with Forward Process](https://mlsanity.com/paper/2509.16117/) (2025)
- [VisionReward: Fine-Grained Multi-Dimensional Human Preference Learning for Image and Video Generation](https://mlsanity.com/paper/2412.21059/) (2024)
- [AutomaTikZ: Text-Guided Synthesis of Scientific Vector Graphics with TikZ](https://mlsanity.com/paper/2310.00367/) (2023)
- [GRPO-Guard: Mitigating Implicit Over-Optimization in Flow Matching via Regulated Clipping](https://mlsanity.com/paper/2510.22319/) (2025)
- [Rendering-Aware Reinforcement Learning for Vector Graphics Generation](https://mlsanity.com/paper/2505.20793/) (2025)
- [From Words to Structured Visuals: A Benchmark and Framework for Text-to-Diagram Generation and Editing](https://mlsanity.com/paper/2411.11916/) (2024)
- [DiagrammerGPT: Generating Open-Domain, Open-Platform Diagrams via LLM Planning](https://mlsanity.com/paper/2310.12128/) (2023)
- [PaperBanana: Automating Academic Illustration for AI Scientists](https://mlsanity.com/paper/2601.23265/) (2026)
- [TikZero: Zero-Shot Text-Guided Graphics Program Synthesis](https://mlsanity.com/paper/2503.11509/) (2025)
- [AutoFigure: Generating and Refining Publication-Ready Scientific Illustrations](https://mlsanity.com/paper/2602.03828/) (2026)
- [FigGen: Text to Scientific Figure Generation](https://mlsanity.com/paper/2306.00800/) (2023)
- [MMMG: A Massive, Multidisciplinary, Multi-Tier Generation Benchmark for Text-to-Image Reasoning](https://mlsanity.com/paper/2506.10963/) (2025)
- [Neighbor GRPO: Contrastive ODE Policy Optimization Aligns Flow Models](https://mlsanity.com/paper/2511.16955/) (2025)
- [SciDoc2Diagrammer-MAF: Towards Generation of Scientific Diagrams from Documents guided by Multi-Aspect Feedback Refinement](https://mlsanity.com/paper/2409.19242/) (2024)
- [PhysCorr: Dual-Reward DPO for Physics-Constrained Text-to-Video Generation with Automated Preference Selection](https://mlsanity.com/paper/2511.03997/) (2025)
- [AutoFigure-Edit: Generating Editable Scientific Illustration](https://mlsanity.com/paper/2603.06674/) (2026)
- [Smart-GRPO: Smartly Sampling Noise for Efficient RL of Flow-Matching Models](https://mlsanity.com/paper/2510.02654/) (2025)
- [AeSlides: Incentivizing Aesthetic Layout in LLM-Based Slide Generation via Verifiable Rewards](https://mlsanity.com/paper/2604.22840/) (2026)
- [GDRO: Group-level Reward Post-training Suitable for Diffusion Models](https://mlsanity.com/paper/2601.02036/) (2026)
- [Paper2SysArch: Structure-Constrained System Architecture Generation from Scientific Papers](https://mlsanity.com/paper/2511.18036/) (2025)
- [Scientific Graphics Program Synthesis via Dual Self-Consistency Reinforcement Learning](https://mlsanity.com/paper/2604.06079/) (2026)
- [SciFlow-Bench: Evaluating Structure-Aware Scientific Diagram Generation via Inverse Parsing](https://mlsanity.com/paper/2602.09809/) (2026)
- [Visual Preference Optimization with Rubric Rewards](https://mlsanity.com/paper/2604.13029/) (2026)
- [AIBench: Evaluating Visual-Logical Consistency in Academic Illustration Generation](https://mlsanity.com/paper/2603.28068/) (2026)
- [GENFIG1: Visual Summaries of Scholarly Work as a Challenge for Vision-Language Models](https://mlsanity.com/paper/2604.04172/) (2026)

## Built on by (most cited first, 1 of 1)

- [Mage-Flow: An Efficient Native-Resolution Foundation Model for Image Generation and Editing](https://mlsanity.com/paper/2607.19064/) (2026)

## Abstract (arXiv)

Structural fidelity is essential to scientific methodology diagrams. To communicate research logic, these diagrams must faithfully render components, directional relations, and textual annotations. Since a single error, such as a reversed arrow or an unreadable equation, can invalidate the entire figure, structural fidelity is inherently conjunctive: correctness on one axis cannot compensate for failure on another. Current open-source models fail to satisfy this criterion. Supervised fine-tuning (SFT) learns plausible layouts but cannot reliably ensure structural correctness, while scalar reward-based post-training obscures which structural dimension has failed. To address this, we introduce SciForma, a framework for the structure faithful generation of scientific methodology diagrams. Specifically, SciForma decomposes diagram quality into three structural axes: Component, Arrow, and Text, guided by a structural inventory. Built on this foundation, we curate SciFormaData-700K for structured training and SciFormaBench-2K for logic-verified evaluation. To close the gap left by SFT, we develop Multi-Dimensional Conjunctive Preference Optimization (M-DPO), which enforces simultaneous correctness across all axes and adaptively routes gradients to the most deficient dimension in post-training. The same structural inventory also enables iterative editing at inference time to correct residual errors. This combination allows SciForma-9B to exceed all open-source baselines and GPT-Image-1.5 on both SciFormaBench-2K and AIBench, bringing open scientific diagram generation close to proprietary-level structural fidelity. Our code and data will be available at: https://github.com/microsoft/SciForma.