# Interpreting Attention Heads for Image-to-Text Information Flow in Large Vision-Language Models

arXiv: https://arxiv.org/abs/2509.17588 · HTML: https://mlsanity.com/paper/2509.17588/

## 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 52 of its 65 references that are in the map.
- First posted on arXiv 22 Sep 2025.

## 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 52)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Layer Normalization](https://mlsanity.com/paper/1607.06450/) (2016)
- [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)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [Efficient Streaming Language Models with Attention Sinks](https://mlsanity.com/paper/2309.17453/) (2023)
- [Evaluating Object Hallucination in Large Vision-Language Models](https://mlsanity.com/paper/2305.10355/) (2023)
- [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://mlsanity.com/paper/2412.05271/) (2024)
- [Root Mean Square Layer Normalization](https://mlsanity.com/paper/1910.07467/) (2019)
- [Moment Matching for Multi-Source Domain Adaptation](https://mlsanity.com/paper/1812.01754/) (2018)
- [Interpretability in the Wild: a Circuit for Indirect Object Identification in GPT-2 small](https://mlsanity.com/paper/2211.00593/) (2022)
- [Toy Models of Superposition](https://mlsanity.com/paper/2209.10652/) (2022)
- [Attention is not Explanation](https://mlsanity.com/paper/1902.10186/) (2019)
- [Are We on the Right Way for Evaluating Large Vision-Language Models?](https://mlsanity.com/paper/2403.20330/) (2024)
- [Vision Transformers Need Registers](https://mlsanity.com/paper/2309.16588/) (2023)
- [Object Hallucination in Image Captioning](https://mlsanity.com/paper/1809.02156/) (2018)
- [An Image is Worth 1/2 Tokens After Layer 2: Plug-and-Play Inference Acceleration for Large Vision-Language Models](https://mlsanity.com/paper/2403.06764/) (2024)
- [Dissecting Recall of Factual Associations in Auto-Regressive Language Models](https://mlsanity.com/paper/2304.14767/) (2023)
- [Sparse Feature Circuits: Discovering and Editing Interpretable Causal Graphs in Language Models](https://mlsanity.com/paper/2403.19647/) (2024)
- [Causal Abstractions of Neural Networks](https://mlsanity.com/paper/2106.02997/) (2021)
- [Mechanistic Interpretability for AI Safety -- A Review](https://mlsanity.com/paper/2404.14082/) (2024)
- [SparseVLM: Visual Token Sparsification for Efficient Vision-Language Model Inference](https://mlsanity.com/paper/2410.04417/) (2024)
- [VisionZip: Longer is Better but Not Necessary in Vision Language Models](https://mlsanity.com/paper/2412.04467/) (2024)
- [PyramidDrop: Accelerating Your Large Vision-Language Models via Pyramid Visual Redundancy Reduction](https://mlsanity.com/paper/2410.17247/) (2024)
- [Label Words are Anchors: An Information Flow Perspective for Understanding In-Context Learning](https://mlsanity.com/paper/2305.14160/) (2023)
- [How to use and interpret activation patching](https://mlsanity.com/paper/2404.15255/) (2024)
- [How does GPT-2 compute greater-than?: Interpreting mathematical abilities in a pre-trained language model](https://mlsanity.com/paper/2305.00586/) (2023)
- [The Hydra Effect: Emergent Self-repair in Language Model Computations](https://mlsanity.com/paper/2307.15771/) (2023)

## Built on by (most cited first, 9 of 9)

- [Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models](https://mlsanity.com/paper/2601.14004/) (2026)
- [Pathways of Visual Information Flow in Vision-Language Models](https://mlsanity.com/paper/2607.03358/) (2026)
- [When Sinks Help or Hurt: Unified Framework for Attention Sink in Large Vision-Language Models](https://mlsanity.com/paper/2604.03316/) (2026)
- [Multimodal LLMs for Historical Dataset Construction from Archival Image Scans: German Patents (1877-1918)](https://mlsanity.com/paper/2512.19675/) (2025)
- [Interpretable Motion-Attentive Maps: Spatio-Temporally Localizing Concepts in Video Diffusion Transformers](https://mlsanity.com/paper/2603.02919/) (2026)
- [MLLMs Know When Before Speaking: Revealing and Recovering Temporal Grounding via Attention Cues](https://mlsanity.com/paper/2605.21954/) (2026)
- [Ego: Embedding-Guided Personalization of Vision-Language Models](https://mlsanity.com/paper/2603.09771/) (2026)
- [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)
- [Understanding Unified VLMs: From Attention Probing to Guidance Tuning for Multimodal Generation](https://mlsanity.com/paper/2610.15227/) (2026)

## Abstract (arXiv)

Large Vision-Language Models (LVLMs) answer visual questions by transferring information from images to text through a series of attention heads. While this image-to-text information flow is central to visual question answering, its underlying mechanism remains difficult to interpret due to the simultaneous operation of numerous attention heads. To address this challenge, we propose head attribution, a technique inspired by component attribution methods, to identify consistent patterns among attention heads that play a key role in information transfer. Using head attribution, we investigate how LVLMs rely on specific attention heads to identify and answer questions about the main object in an image. Our analysis reveals that a distinct subset of attention heads facilitates the image-to-text information flow. Remarkably, we find that the selection of these heads is governed by the semantic content of the input image rather than its visual appearance. We further examine the flow of information at the token level and discover that (1) text information first propagates to role-related tokens and the final token before receiving image information, and (2) image information is embedded in both object-related and background tokens. Our work provides evidence that image-to-text information flow follows a structured process, and that analysis at the attention-head level offers a promising direction toward understanding the mechanisms of LVLMs.