mlsanityData to 14 Oct 2026

From Behavioral Performance to Internal Competence: Interpreting Vision-Language Models with VLM-Lens

arXiv 2510.02292 ↗First posted 2 Oct 2025cs.CL cs.CVcited by 9

On the map: Multimodal LLMs / vision-language models → VLM benchmarks and visual reasoning

Abstract (arXiv)

We introduce VLM-Lens, a toolkit designed to enable systematic benchmarking, analysis, and interpretation of vision-language models (VLMs) by supporting the extraction of intermediate outputs from any layer during the forward pass of open-source VLMs. VLM-Lens provides a unified, YAML-configurable interface that abstracts away model-specific complexities and supports user-friendly operation across diverse VLMs. It currently supports 16 state-of-the-art base VLMs and their over 30 variants, and is extensible to accommodate new models without changing the core logic. The toolkit integrates easily with various interpretability and analysis methods. We demonstrate its usage with two simple analytical experiments, revealing systematic differences in the hidden representations of VLMs across layers and target concepts. VLM-Lens is released as an open-sourced project to accelerate community efforts in understanding and improving VLMs.

Key facts. Cited by 9 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 27 of its 47 references that are in the map. First posted on arXiv 2 Oct 2025.

Builds on

27 of 47 references.

Closest ancestors · 1

Same direction, newest first

  1. 2023GLaMM: Pixel Grounding Large Multimodal Model6 Nov

Other references · 26

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2020Language Models are Few-Shot Learnerscited by 32,419
  3. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,305
  4. 2014Microsoft COCO: Common Objects in Contextcited by 17,608
  5. 2023Segment Anythingcited by 8,234
  6. 2016Grad-CAM: Visual Explanations from Deep Networks via Gradient-based Localizationcited by 6,733
  7. 2016Layer Normalizationcited by 5,669
  8. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  9. 2023InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Taskscited by 2,565
  10. 2023MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGIcited by 2,080
  11. 2016CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoningcited by 1,457
  12. 2024MiniCPM-V: A GPT-4V Level MLLM on Your Phonecited by 1,079
14 more
  1. 2019What Does BERT Look At? An Analysis of BERT's Attentioncited by 1,003
  2. 2019Designing and Interpreting Probes with Control Taskscited by 640
  3. 2023Eliciting Latent Predictions from Transformers with the Tuned Lenscited by 623
  4. 2022Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionalitycited by 529
  5. 2023InternLM-XComposer: A Vision-Language Large Model for Advanced Text-image Comprehension and Compositioncited by 297
  6. 2014Unifying Visual-Semantic Embeddings with Multimodal Neural Language Modelscited by 198
  7. 2023A Toy Model of Universality: Reverse Engineering How Networks Learn Group Operationscited by 144
  8. 2025PerceptionLM: Open-Access Data and Models for Detailed Visual Understandingcited by 92
  9. 2023Towards Vision-Language Mechanistic Interpretability: A Causal Tracing Tool for BLIPcited by 62
  10. 2024Do Vision-Language Models Represent Space and How? Evaluating Spatial Frame of Reference Under Ambiguitiescited by 59
  11. 2025Aya Vision: Advancing the Frontier of Multilingual Multimodalitycited by 33
  12. 2025Do Vision-Language Models Have Internal World Models? Towards an Atomic Evaluationcited by 24
  13. 2024Large Multi-modal Models Can Interpret Features in Large Multi-modal Modelscited by 23
  14. 2025Prisma: An Open Source Toolkit for Mechanistic Interpretability in Vision and Videocited by 21

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 3; 2026: 6.

9 papers.

  1. 2026HALP: Detecting Hallucinations in Vision-Language Models without Generating a Single Tokencited by 14
  2. 2026Seeing without Looking: Do Vision-Language Benchmarks Really Test Vision?cited by 6
  3. 2025Video Models Start to Solve Chess, Maze, Sudoku, Mental Rotation, and Raven' Matricescited by 4
  4. 2025When to Think and When to Look: Uncertainty-Guided Lookbackcited by 3
  5. 2026Why Far Looks Up: Probing Spatial Representation in Vision-Language Modelscited by 3
  6. 2026Where Vision Becomes Text: Locating the OCR Routing Bottleneck in Vision-Language Modelscited by 2
  7. 2025What Color Is the Text? A Benchmark for Hallucination Induced by Image-Embedded Promptcited by 1
  8. 2026Evaluating Reasoning Fidelity in Visual Text Generationcited by 0
  9. 2026When Safety Overrides Vision: Exploring Dynamics between Vision Influence and Safety Alignment in Vision-Language Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024VHELM: A Holistic Evaluation of Vision Language Models
  2. 2026DiaVLo: Diagnosing Behaviours of Vision-Language Models
  3. 2025Hidden in plain sight: VLMs overlook their visual representations
  4. 2026Empirical Recipes for Efficient and Compact Vision-Language Models
  5. 2026JAMMEval: A Refined Collection of Japanese Benchmarks for Reliable VLM Evaluation
  6. 2025Bridging vision language model (VLM) evaluation gaps with a framework for scalable and cost-effective benchmark generation
  7. 2024What matters when building vision-language models?
  8. 2026PRISM-VLM: A Multi-Axis Discriminative Benchmark for Compact Vision-Language Models