# Vision Function Layer in Multimodal LLMs

arXiv: https://arxiv.org/abs/2509.24791 · HTML: https://mlsanity.com/paper/2509.24791/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 61 of its 72 references that are in the map.
- First posted on arXiv 29 Sep 2025.

## Placed in

- Multimodal LLMs / vision-language models / [MLLM training and visual instruction tuning](https://mlsanity.com/mllm/mllm-training-and-visual-instruction-tuning/)

## Builds on (most cited first, 30 of 61)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale](https://mlsanity.com/paper/2010.11929/) (2020)
- [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)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models](https://mlsanity.com/paper/2301.12597/) (2023)
- [PaLM: Scaling Language Modeling with Pathways](https://mlsanity.com/paper/2204.02311/) (2022)
- [Qwen2.5 Technical Report](https://mlsanity.com/paper/2412.15115/) (2024)
- [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)
- [SQuAD: 100,000+ Questions for Machine Comprehension of Text](https://mlsanity.com/paper/1606.05250/) (2016)
- [Qwen Technical Report](https://mlsanity.com/paper/2309.16609/) (2023)
- [Training Compute-Optimal Large Language Models](https://mlsanity.com/paper/2203.15556/) (2022)
- [LLaVA-OneVision: Easy Visual Task Transfer](https://mlsanity.com/paper/2408.03326/) (2024)
- [InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks](https://mlsanity.com/paper/2312.14238/) (2023)
- [Character-level Convolutional Networks for Text Classification](https://mlsanity.com/paper/1509.01626/) (2015)
- [MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI](https://mlsanity.com/paper/2311.16502/) (2023)
- [BLOOM: A 176B-Parameter Open-Access Multilingual Language Model](https://mlsanity.com/paper/2211.05100/) (2022)
- [Learn to Explain: Multimodal Reasoning via Thought Chains for Science Question Answering](https://mlsanity.com/paper/2209.09513/) (2022)
- [CommonsenseQA: A Question Answering Challenge Targeting Commonsense Knowledge](https://mlsanity.com/paper/1811.00937/) (2018)
- [Evaluating Object Hallucination in Large Vision-Language Models](https://mlsanity.com/paper/2305.10355/) (2023)
- [Towards VQA Models That Can Read](https://mlsanity.com/paper/1904.08920/) (2019)
- [MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models](https://mlsanity.com/paper/2306.13394/) (2023)
- [Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scaling](https://mlsanity.com/paper/2412.05271/) (2024)

## Built on by (most cited first, 17 of 17)

- [Demystifying Video Reasoning](https://mlsanity.com/paper/2603.16870/) (2026)
- [VGent: Visual Grounding via Modular Design for Disentangling Reasoning and Prediction](https://mlsanity.com/paper/2512.11099/) (2025)
- [Q-Zoom: Query-Aware Adaptive Perception for Efficient Multimodal Large Language Models](https://mlsanity.com/paper/2604.06912/) (2026)
- [What does RL improve for Visual Reasoning? A Frankenstein-Style Analysis](https://mlsanity.com/paper/2602.12395/) (2026)
- [When Sinks Help or Hurt: Unified Framework for Attention Sink in Large Vision-Language Models](https://mlsanity.com/paper/2604.03316/) (2026)
- [PlaM: Training-Free Plateau-Guided Model Merging for Better Visual Grounding in MLLMs](https://mlsanity.com/paper/2601.07645/) (2026)
- [RefAny3D: 3D Asset-Referenced Diffusion Models for Image Generation](https://mlsanity.com/paper/2601.22094/) (2026)
- [Look Twice: Training-Free Evidence Highlighting for Knowledge-based Visual Question Answering](https://mlsanity.com/paper/2604.01280/) (2026)
- [Do Vision Language Models Need to Process Image Tokens?](https://mlsanity.com/paper/2604.09425/) (2026)
- [ERASE: Eliminating Redundant Visual Tokens via Adaptive Two-Stage Token Pruning](https://mlsanity.com/paper/2605.09982/) (2026)
- [When Structured Sparse Autoencoders Learn Consistent Concepts Across Modalities](https://mlsanity.com/paper/2607.08605/) (2026)
- [System-Mediated Attention Imbalances Make Vision-Language Models Say Yes](https://mlsanity.com/paper/2601.12430/) (2026)
- [ConFoThinking: Consolidated Focused Attention Driven Thinking for Visual Question Answering](https://mlsanity.com/paper/2603.00165/) (2026)
- [Don't Look at the Numbers: Visual Anchoring Bias and Layer-wise Representation in VLMs](https://mlsanity.com/paper/2605.11218/) (2026)
- [LAVE: Latent Visual Evidence-Enhanced Planning for Video Tool-use Agents](https://mlsanity.com/paper/2608.07585/) (2026)
- [Layer-Aware Position Embeddings for Visual Token Pruning in Multimodal Large Language Models](https://mlsanity.com/paper/2609.23715/) (2026)
- [The Evidence Anchoring Boundary: Where Visual Interventions Matter and When They Help in Multimodal LLMs](https://mlsanity.com/paper/2610.12981/) (2026)

## Abstract (arXiv)

This study identifies that visual-related functional decoding is distributed across different decoder layers in Multimodal Large Language Models (MLLMs). Typically, each function, such as counting, grounding, or OCR recognition, narrows down to two or three layers, which we define as Vision Function Layers (VFL). Additionally, the depth and its order of different VFLs exhibits a consistent pattern across different MLLMs, which is well-aligned with human behaviors (e.g., recognition occurs first, followed by counting, and then grounding). These findings are derived from Visual Token Swapping, our novel analytical framework that modifies targeted KV cache entries to precisely elucidate layer-specific functions during decoding. Furthermore, these insights offer substantial utility in tailoring MLLMs for real-world downstream applications. For instance, when LoRA training is selectively applied to VFLs whose functions align with the training data, VFL-LoRA not only outperform full-LoRA but also prevent out-of-domain function forgetting. Moreover, by analyzing the performance differential on training data when particular VFLs are ablated, VFL-select automatically classifies data by function, enabling highly efficient data selection to directly bolster corresponding capabilities. Consequently, VFL-select surpasses human experts in data selection, and achieves 98% of full-data performance with only 20% of the original dataset. This study delivers deeper comprehension of MLLM visual processing, fostering the creation of more efficient, interpretable, and robust models.