# How Do VLMs Behave When Blind or Misled? Behavioral Evaluation of VLMs on Scientific Figures

arXiv: https://arxiv.org/abs/2608.13267 · HTML: https://mlsanity.com/paper/2608.13267/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Multimodal reasoning, one of 17 directions in LLM reasoning.
- Builds on 20 of its 45 references that are in the map.
- First posted on arXiv 13 Aug 2026.

## Placed in

- LLM reasoning / [Multimodal reasoning](https://mlsanity.com/lr/multimodal-reasoning/)
- Multimodal LLMs / vision-language models / [VLM benchmarks and visual reasoning](https://mlsanity.com/mllm/vlm-benchmarks-and-visual-reasoning/)

## Builds on (most cited first, 20 of 20)

- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI](https://mlsanity.com/paper/2311.16502/) (2023)
- [Language Models (Mostly) Know What They Know](https://mlsanity.com/paper/2207.05221/) (2022)
- [MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts](https://mlsanity.com/paper/2310.02255/) (2023)
- [ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning](https://mlsanity.com/paper/2203.10244/) (2022)
- [Just Ask for Calibration: Strategies for Eliciting Calibrated Confidence Scores from Language Models Fine-Tuned with Human Feedback](https://mlsanity.com/paper/2305.14975/) (2023)
- [Phi-4-Mini Technical Report: Compact yet Powerful Multimodal Language Models via Mixture-of-LoRAs](https://mlsanity.com/paper/2503.01743/) (2025)
- [Experts, Errors, and Context: A Large-Scale Study of Human Evaluation for Machine Translation](https://mlsanity.com/paper/2104.14478/) (2021)
- [ChartBench: A Benchmark for Complex Visual Reasoning in Charts](https://mlsanity.com/paper/2312.15915/) (2023)
- [Do LVLMs Understand Charts? Analyzing and Correcting Factual Errors in Chart Captioning](https://mlsanity.com/paper/2312.10160/) (2023)
- [SciFIBench: Benchmarking Large Multimodal Models for Scientific Figure Interpretation](https://mlsanity.com/paper/2405.08807/) (2024)
- [Selective "Selective Prediction": Reducing Unnecessary Abstention in Vision-Language Reasoning](https://mlsanity.com/paper/2402.15610/) (2024)
- [A Survey of Scientific Large Language Models: From Data Foundations to Agent Frontiers](https://mlsanity.com/paper/2508.21148/) (2025)
- [MultiChartQA: Benchmarking Vision-Language Models on Multi-Chart Problems](https://mlsanity.com/paper/2410.14179/) (2024)
- [ScImage: How Good Are Multimodal Large Language Models at Scientific Text-to-Image Generation?](https://mlsanity.com/paper/2412.02368/) (2024)
- [EncQA: Benchmarking Vision-Language Models on Visual Encodings for Charts](https://mlsanity.com/paper/2508.04650/) (2025)
- [Protecting multimodal large language models against misleading visualizations](https://mlsanity.com/paper/2502.20503/) (2025)
- [TikZilla: Scaling Text-to-TikZ with High-Quality Data and Reinforcement Learning](https://mlsanity.com/paper/2603.03072/) (2026)
- [ChartHal: A Fine-grained Framework Evaluating Hallucination of Large Vision Language Models in Chart Understanding](https://mlsanity.com/paper/2509.17481/) (2025)
- [CHAOS: Chart Analysis with Outlier Samples](https://mlsanity.com/paper/2505.17235/) (2025)

## Abstract (arXiv)

Existing vision-language model (VLM) benchmarks emphasize perception and reasoning accuracy (how well VLMs describe and reason about what they see in an image), with limited attention to behavioral reliability under uncertainty (how they behave when visual evidence is missing or misleading). We introduce SciFigBench, a diagnostic VLM benchmark for scientific figure understanding that jointly evaluates perception, reasoning, and behavioral reliability under uncertainty. It contains 250 figures with high-quality human annotations across three evaluation aspects, totaling 600+ hours of annotation effort. We further extend these figures via image transformations, reasoning questions, resistance probes, caption-bias probes, and confirmed selective-blur targets, producing over 34,000 evaluation setups for stress testing. We further propose the Admittance-Resistance-Inductance (A-R-I) framework to evaluate whether models acknowledge insufficient evidence, resist misleading context, and infer cautiously from partial information. Our results reveal substantial behavioral differences among models. GPT-5.2 achieves the highest description quality (MQM 91.6) with strong reasoning accuracy (78.4%), yet hallucinates unreadable content in 96% of cases, whereas Gemini 3.1 Pro, a comparably capable model (MQM 90.2, reasoning 81.0%), admits uncertainty in 71% of such cases and achieves the strongest resistance score (0.91). These findings show that high perception and reasoning accuracy alone do not guarantee behavioral reliability, a dimension critical for deployment in scientific workflows.