# VLM benchmarks and visual reasoning Area: [Multimodal LLMs / vision-language models](https://mlsanity.com/mllm/index.md). HTML: https://mlsanity.com/mllm/vlm-benchmarks-and-visual-reasoning/ ## Key facts - VLM benchmarks and visual reasoning: 2,063 papers on arXiv since 2021 (data through 14 Oct 2026). - 265 of its 2,063 mapped papers (13%) cite “Flamingo: a Visual Language Model for Few-Shot Learning” (2022, arXiv 2204.14198). - 256 of its 2,063 mapped papers (12%) cite “MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts” (2023, arXiv 2310.02255). - 89 new papers 17 Sep–14 Oct. ## Most cited here - [Flamingo: a Visual Language Model for Few-Shot Learning](https://mlsanity.com/paper/2204.14198/) (2022), cited by 265 papers here - [MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts](https://mlsanity.com/paper/2310.02255/) (2023), cited by 256 papers here - [Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond](https://mlsanity.com/paper/2308.12966/) (2023), cited by 227 papers here - [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://mlsanity.com/paper/2201.12086/) (2022), cited by 199 papers here - [Are We on the Right Way for Evaluating Large Vision-Language Models?](https://mlsanity.com/paper/2403.20330/) (2024), cited by 146 papers here - [Measuring Multimodal Mathematical Reasoning with MATH-Vision Dataset](https://mlsanity.com/paper/2402.14804/) (2024), cited by 120 papers here - [MathVerse: Does Your Multi-modal LLM Truly See the Diagrams in Visual Math Problems?](https://mlsanity.com/paper/2403.14624/) (2024), cited by 114 papers here - [Kosmos-2: Grounding Multimodal Large Language Models to the World](https://mlsanity.com/paper/2306.14824/) (2023), cited by 96 papers here - [Visual CoT: Advancing Multi-Modal Language Models with a Comprehensive Dataset and Benchmark for Chain-of-Thought Reasoning](https://mlsanity.com/paper/2403.16999/) (2024), cited by 93 papers here - [LLaVA-CoT: Let Vision Language Models Reason Step-by-Step](https://mlsanity.com/paper/2411.10440/) (2024), cited by 86 papers here ## New (17 Sep–14 Oct) - [Probing Negation Understanding in Vision-Language Models With Object Attention](https://mlsanity.com/paper/2610.20852/) (2026-10-14) - [ReferFlow: Beyond Final-Target Localization for Complex Referring Expressions](https://mlsanity.com/paper/2610.20352/) (2026-10-14) - [VisualWorldBench: A Fine-Grained Multi-Task Benchmark for Evaluating Visual World Knowledge in MLLMs](https://mlsanity.com/paper/2610.20083/) (2026-10-14) - [Segment Where You Look: Attention as the Interface Between Language Models and Segmenters](https://mlsanity.com/paper/2610.18729/) (2026-10-13) - [CREATE: Creativity Evaluation of VLMs using the Visual Remote Associates Test](https://mlsanity.com/paper/2610.18569/) (2026-10-13) - [Seeing Depth but Choosing Cues in Vision-Language Models](https://mlsanity.com/paper/2610.17976/) (2026-10-13) - [The Value of Being Wrong: Self-Mined Visual In-Context Learning from Errors](https://mlsanity.com/paper/2610.16977/) (2026-10-12) - [Teaching Vision-Language Models to Decline: Support Completion Against Mirage](https://mlsanity.com/paper/2610.16465/) (2026-10-12) - [Adaptive Look, Flexible Thought: Learning How and How Long to Think in Multimodal Reasoning](https://mlsanity.com/paper/2610.15810/) (2026-10-11) - [Genesis: Visual Augmentation Primitives for Multimodal Chain-of-Thought Reasoning](https://mlsanity.com/paper/2610.15565/) (2026-10-11) - [Semantic Saliency from Multi-Modal Large Language Model Scene Understanding Maps](https://mlsanity.com/paper/2610.15467/) (2026-10-11) - [Understanding Unified VLMs: From Attention Probing to Guidance Tuning for Multimodal Generation](https://mlsanity.com/paper/2610.15227/) (2026-10-10) - [Thinking with Images without Coordinates](https://mlsanity.com/paper/2610.14265/) (2026-10-09) - [Cross-Modal Latent Influence: Patch-Level Training Sample Attribution for Vision-Language Models](https://mlsanity.com/paper/2610.14190/) (2026-10-09) - [When VLMs Destroy What They Know: Spatial Representations Fall Below Lexical Priors](https://mlsanity.com/paper/2610.13910/) (2026-10-09) - [OccludeBench: Can VLMs Find What's Missing?](https://mlsanity.com/paper/2610.13790/) (2026-10-09) - [Auditing VLM Benchmarks from Existing Evaluation Runs](https://mlsanity.com/paper/2610.13548/) (2026-10-09) - [When MLLMs miss what humans read: diagnosing and learning to read squint text](https://mlsanity.com/paper/2610.13465/) (2026-10-09) - [MHR-Router: Multimodal Heterogeneous Retrieval-Enhanced Adaptive Routing for VLM Inference](https://mlsanity.com/paper/2610.13278/) (2026-10-09) - [What Do Visual-Head Discovery Scores Actually Find? Attention Source, Necessity, and Random-Control Pitfalls in Vision-Language Models](https://mlsanity.com/paper/2610.13111/) (2026-10-09) - [MVGraph: Benchmarking LVLMs on Evidence Composition across Multi-View Visual Graph](https://mlsanity.com/paper/2610.12961/) (2026-10-09) - [In-Context Learning Can Help Vision-Language Models Overcome Training Priors](https://mlsanity.com/paper/2610.12942/) (2026-10-09) - [From What to Which: Decoding Modifier Grounding in Frozen MLLMs](https://mlsanity.com/paper/2610.12305/) (2026-10-08) - [Harness Compilation: Which Decisions Should a Small Vision-Language Model Keep?](https://mlsanity.com/paper/2610.11231/) (2026-10-08) - [Visual Evidence Under Cross-Examination: Evaluating and Controlling Decision-Level Evidence Use in Vision-Language Models](https://mlsanity.com/paper/2610.09550/) (2026-10-07) - [GeoPID: Decomposing and Steering Visual Information in Vision-Language Models](https://mlsanity.com/paper/2610.08401/) (2026-10-06) - [Matching Object or Relation? Tracing Abstract Reasoning Inside VLMs](https://mlsanity.com/paper/2610.07646/) (2026-10-06) - [Harnessing Multimodal Large Language Models for Training-Free Human-Object Interaction Detection](https://mlsanity.com/paper/2610.06394/) (2026-10-05) - [Readout Blindness: VLM Scores Miss the Spatial Direction Their Frozen Encoders Retain](https://mlsanity.com/paper/2610.06324/) (2026-10-05) - [When to Rethink: Learning Multi-Perspective Self-Verification for Vision-Language Models](https://mlsanity.com/paper/2610.07018/) (2026-10-04)