# NaturalBench: Evaluating Vision-Language Models on Natural Adversarial Samples

arXiv: https://arxiv.org/abs/2410.14669 · HTML: https://mlsanity.com/paper/2410.14669/

## Key facts

- Cited by 97 papers in the map (data through 14 Oct 2026).
- Placed in VLM benchmarks and visual reasoning, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 70 of its 93 references that are in the map.
- First posted on arXiv 18 Oct 2024.

## Placed in

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

## Builds on (most cited first, 30 of 70)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Visual Instruction Tuning](https://mlsanity.com/paper/2304.08485/) (2023)
- [Explaining and Harnessing Adversarial Examples](https://mlsanity.com/paper/1412.6572/) (2014)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [Improved Baselines with Visual Instruction Tuning](https://mlsanity.com/paper/2310.03744/) (2023)
- [Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution](https://mlsanity.com/paper/2409.12191/) (2024)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning](https://mlsanity.com/paper/2305.06500/) (2023)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [Visual Genome: Connecting Language and Vision Using Crowdsourced Dense Image Annotations](https://mlsanity.com/paper/1602.07332/) (2016)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering](https://mlsanity.com/paper/1612.00837/) (2016)
- [VQA: Visual Question Answering](https://mlsanity.com/paper/1505.00468/) (2015)
- [MMBench: Is Your Multi-modal Model an All-around Player?](https://mlsanity.com/paper/2307.06281/) (2023)
- [MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI](https://mlsanity.com/paper/2311.16502/) (2023)
- [Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering](https://mlsanity.com/paper/2209.09513/) (2022)
- [A Survey on Bias and Fairness in Machine Learning](https://mlsanity.com/paper/1908.09635/) (2019)
- [Phi-3 Technical Report: A Highly Capable Language Model Locally on Your Phone](https://mlsanity.com/paper/2404.14219/) (2024)
- [MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models](https://mlsanity.com/paper/2306.13394/) (2023)
- [MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts](https://mlsanity.com/paper/2310.02255/) (2023)
- [Natural Adversarial Examples](https://mlsanity.com/paper/1907.07174/) (2019)
- [Flickr30k Entities: Collecting Region-to-Phrase Correspondences for Richer Image-to-Sentence Models](https://mlsanity.com/paper/1505.04870/) (2015)
- [MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities](https://mlsanity.com/paper/2308.02490/) (2023)
- [How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites](https://mlsanity.com/paper/2404.16821/) (2024)
- [mPLUG-Owl: Modularization Empowers Large Language Models with Multimodality](https://mlsanity.com/paper/2304.14178/) (2023)
- [ShareGPT4V: Improving Large Multi-Modal Models with Better Captions](https://mlsanity.com/paper/2311.12793/) (2023)

## Built on by (most cited first, 30 of 97)

- [VL-Uncertainty: Detecting Hallucination in Large Vision-Language Model via Uncertainty Estimation](https://mlsanity.com/paper/2411.11919/) (2024)
- [Vision-Language Models Do Not Understand Negation](https://mlsanity.com/paper/2501.09425/) (2025)
- [SimpleVQA: Multimodal Factuality Evaluation for Multimodal Large Language Models](https://mlsanity.com/paper/2502.13059/) (2025)
- [Towards Understanding Camera Motions in Any Video](https://mlsanity.com/paper/2504.15376/) (2025)
- [Grounded Chain-of-Thought for Multimodal Large Language Models](https://mlsanity.com/paper/2503.12799/) (2025)
- [A Survey of State of the Art Large Vision Language Models: Alignment, Benchmark, Evaluations and Challenges](https://mlsanity.com/paper/2501.02189/) (2025)
- [Probing Visual Language Priors in VLMs](https://mlsanity.com/paper/2501.00569/) (2024)
- [Feather the Throttle: Revisiting Visual Token Pruning for Vision-Language Model Acceleration](https://mlsanity.com/paper/2412.13180/) (2024)
- [CAPTURe: Evaluating Spatial Reasoning in Vision Language Models via Occluded Object Counting](https://mlsanity.com/paper/2504.15485/) (2025)
- [A Stitch in Time Saves Nine: Small VLM is a Precise Guidance for Accelerating Large VLMs](https://mlsanity.com/paper/2412.03324/) (2024)
- [VLM2-Bench: A Closer Look at How Well VLMs Implicitly Link Explicit Matching Visual Cues](https://mlsanity.com/paper/2502.12084/) (2025)
- [ICONS: Influence Consensus for Vision-Language Data Selection](https://mlsanity.com/paper/2501.00654/) (2024)
- [ColorBench: Can VLMs See and Understand the Colorful World? A Comprehensive Benchmark for Color Perception, Reasoning, and Robustness](https://mlsanity.com/paper/2504.10514/) (2025)
- [Explain Before You Answer: A Survey on Compositional Visual Reasoning](https://mlsanity.com/paper/2508.17298/) (2025)
- [Unfolding Spatial Cognition: Evaluating Multimodal Models on Visual Simulations](https://mlsanity.com/paper/2506.04633/) (2025)
- [Are They the Same? Exploring Visual Correspondence Shortcomings of Multimodal LLMs](https://mlsanity.com/paper/2501.04670/) (2025)
- [GenHancer: Imperfect Generative Models are Secretly Strong Vision-Centric Enhancers](https://mlsanity.com/paper/2503.19480/) (2025)
- [Turbo3D: Ultra-fast Text-to-3D Generation](https://mlsanity.com/paper/2412.04470/) (2024)
- [Symmetrical Visual Contrastive Optimization: Aligning Vision-Language Models with Minimal Contrastive Images](https://mlsanity.com/paper/2502.13928/) (2025)
- [Escaping the SpuriVerse: Can Large Vision-Language Models Generalize Beyond Seen Spurious Correlations?](https://mlsanity.com/paper/2506.18322/) (2025)
- [VLM-RobustBench: A Comprehensive Benchmark for Robustness of Vision-Language Models](https://mlsanity.com/paper/2603.06148/) (2026)
- [When Semantics Mislead Vision: Mitigating Large Multimodal Models Hallucinations in Scene Text Spotting and Understanding](https://mlsanity.com/paper/2506.05551/) (2025)
- [Exploring How Generative MLLMs Perceive More Than CLIP with the Same Vision Encoder](https://mlsanity.com/paper/2411.05195/) (2024)
- [Enhancing Few-Shot Vision-Language Classification with Large Multimodal Model Features](https://mlsanity.com/paper/2412.00142/) (2024)
- [TimeBlind: A Spatio-Temporal Compositionality Benchmark for Video LLMs](https://mlsanity.com/paper/2602.00288/) (2026)
- [MultiVerse: A Multi-Turn Conversation Benchmark for Evaluating Large Vision and Language Models](https://mlsanity.com/paper/2510.16641/) (2025)
- [BLINK-Twice: You see, but do you observe? A Reasoning Benchmark on Visual Perception](https://mlsanity.com/paper/2510.09361/) (2025)
- [Res-Bench: Benchmarking the Robustness of Multimodal Large Language Models to Dynamic Resolution Input](https://mlsanity.com/paper/2510.16926/) (2025)
- [UniGame: Turning a Unified Multimodal Model Into Its Own Adversary](https://mlsanity.com/paper/2511.19413/) (2025)
- [From Easy to Hard: The MIR Benchmark for Progressive Interleaved Multi-Image Reasoning](https://mlsanity.com/paper/2509.17040/) (2025)

## Abstract (arXiv)

Vision-language models (VLMs) have made significant progress in recent visual-question-answering (VQA) benchmarks that evaluate complex visio-linguistic reasoning. However, are these models truly effective? In this work, we show that VLMs still struggle with natural images and questions that humans can easily answer, which we term natural adversarial samples. We also find it surprisingly easy to generate these VQA samples from natural image-text corpora using off-the-shelf models like CLIP and ChatGPT. We propose a semi-automated approach to collect a new benchmark, NaturalBench, for reliably evaluating VLMs with 10,000 human-verified VQA samples. Crucially, we adopt a vision-centric design by pairing each question with two images that yield different answers, preventing blind solutions from answering without using the images. This makes NaturalBench more challenging than previous benchmarks that can be solved with commonsense priors. We evaluate 53 state-of-the-art VLMs on NaturalBench, showing that models like LLaVA-OneVision, Cambrian-1, Llama3.2-Vision, Molmo, Qwen2-VL, and even GPT-4o lag 50%-70% behind human performance (over 90%). We analyze why NaturalBench is hard from two angles: (1) Compositionality: Solving NaturalBench requires diverse visio-linguistic skills, including understanding attribute bindings, object relationships, and advanced reasoning like logic and counting. To this end, unlike prior work that uses a single tag per sample, we tag each NaturalBench sample with 1 to 8 skill tags for fine-grained evaluation. (2) Biases: NaturalBench exposes severe biases in VLMs, as models often choose the same answer regardless of the image. Lastly, we apply our benchmark curation method to diverse data sources, including long captions (over 100 words) and non-English languages like Chinese and Hindi, highlighting its potential for dynamic evaluations of VLMs.