mlsanityData to 14 Oct 2026

When Looks Do Not Lie: Discourse Structure Guided In-Context Learning for Faithful Diagram Generation

arXiv 2601.20476 ↗First posted 28 Jan 2026cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

GenAI is widespread in educational applications; however, it is known to generate content with intrinsic and extrinsic hallucination. We introduce a novel method for ICL diagram generation based on Rhetorical Structure Theory, which improves diagram faithfulness to its source text context. We find that ICL performance depends on task distribution and models' reasoning ability, with higher reasoning allowing better quality and performance for an out-of-distribution task. We perform an expert evaluation of 150 generated diagrams and analyze our findings using Bayesian GLMMs. Additionally, we use our evaluation rubric and samples from the data set for automated diagram evaluation, achieving statistically significant agreement with human evaluation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 17 of its 92 references that are in the map. First posted on arXiv 28 Jan 2026.

Builds on

17 of 92 references.

References · 17

  1. 2022Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?cited by 1,434
  2. 2021An Explanation of In-context Learning as Implicit Bayesian Inferencecited by 938
  3. 2015Practical Bayesian model evaluation using leave-one-out cross-validation and WAICcited by 110
  4. 2019Rank-normalization, folding, and localization: An improved widehatR for assessing convergence of MCMCcited by 93
  5. 2024Unfamiliar Finetuning Examples Control How Language Models Hallucinatecited by 87
  6. 2022On the Relation between Sensitivity and Accuracy in In-context Learningcited by 72
  7. 2019AI2D-RST: A multimodal corpus of 1000 primary school science diagramscited by 60
  8. 2023In-Context Learning Learns Label Relationships but Is Not Conventional Learningcited by 51
  9. 2025HALoGEN: Fantastic LLM Hallucinations and Where to Find Themcited by 34
  10. 2024Can In-context Learning Really Generalize to Out-of-distribution Tasks?cited by 23
  11. 2025When can in-context learning generalize out of task distribution?cited by 16
  12. 2024SciDoc2Diagrammer-MAF: Towards Generation of Scientific Diagrams from Documents guided by Multi-Aspect Feedback Refinementcited by 15
5 more
  1. 2025No Need for Explanations: LLMs can implicitly learn from mistakes in-contextcited by 13
  2. 2025Where to show Demos in Your Prompt: A Positional Bias of In-Context Learningcited by 13
  3. 2017Discourse Structure in Machine Translation Evaluationcited by 7
  4. 2021krippendorffsalpha: An R Package for Measuring Agreement Using Krippendorff's Alpha Coefficientcited by 7
  5. 2025Beyond the Benefits: A Systematic Review of the Harms and Consequences of Generative AI in Computing Educationcited by 4

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025Draw with Thought: Unleashing Multimodal Reasoning for Scientific Diagram Generation
  2. 2026Diagram-MMU: A Multi-Modal Benchmark for Scientific Diagrams
  3. 2025MAGMA-Edu: Multi-Agent Generative Multimodal Framework for Text-Diagram Educational Question Generation
  4. 2025Structured Visualization Design Knowledge for Grounding Generative Reasoning and Situated Feedback
  5. 2026GenAI-DrawIO-Creator: A Framework for Automated Diagram Generation
  6. 2026SketchJudge: A Diagnostic Benchmark for Grading Hand-drawn Diagrams with Multimodal Large Language Models
  7. 2026From Terminology to Diagrams: Visual-Instruction Generation for Scientific Diagram Understanding
  8. 2026SciForma: Structure-Faithful Generation of Scientific Diagrams