cs.CVMar 18, 2026

From Drop-off to Recovery: A Mechanistic Analysis of Segmentation in MLLMs

Authors: Boyong Wu, Sanghwan Kim, Zeynep Akata

Organizations: Technical University of Munich · Helmholtz Munich · Munich Center for Machine Learning (MCML)

Read PDF
Ask AI
Cite

Copy citation

DOINot available for this paper

Abstract

Multimodal Large Language Models (MLLMs) are increasingly applied to pixel-level vision tasks, yet their intrinsic capacity for spatial understanding remains poorly understood. We investigate segmentation capacity through a layerwise linear probing evaluation across the entire MLLM pipeline: vision encoder, adapter, and LLM. We further conduct an intervention based attention knockout analysis to test whether cross-token attention progressively refines visual representations, and an evaluation of bidirectional attention among image tokens on spatial consistency. Our analysis reveals that the adapter introduces a segmentation representation drop-off, but LLM layers progressively recover through attention-mediated refinement, where correctly classified tokens steer misclassified neighbors toward the correct label. At early image token positions, this recovery is bounded by causal attention, which bidirectional attention among image tokens alleviates. These findings provide a mechanistic account of how MLLMs process visual information for segmentation, informing the design of future segmentation-capable models.

Figures & tables

Figure 2: Overview of the three analysis methods. (a) Layerwise linear probing: given an input image, the vision encoder produces patch token embeddings (brown), which are projected by the adapter into the LLM’s embedding space. Inside the LLM, image tokens are processed jointly with text prompt tokens (yellow). At each layer ℓ\ell , we extract only the image token representations and train an independent linear probe to predict per-patch semantic classes, reassembled into a 2D segmentation map. (b) Attention knockout: we selectively block attention to incorrectly classified tokens (left) or correctly classified tokens (right) across all LLM layers, testing whether cross-token attention drives self-refinement. (c) Bidirectional attention mask: image tokens attend to each other bidirectionally while all other token pairs retain causal masking, alleviating context starvation at early image positions.
Figure 3: Layerwise linear probing results across the MLLM stack. mIoU on ADE20K for CLIP, DINOv2, and SigLIP encoders paired with Vicuna-7B, measured at the vision encoder output, adapter output, and each LLM layer. All three encoders exhibit a drop at the adapter followed by progressive recovery across LLM layers. Dashed lines mark the best-performing layer; values on the right indicate the total mIoU improvement across LLM layers.
Figure 4: Qualitative segmentation predictions across the MLLM stack. From left to right: input image, ground truth, linear probe prediction at the vision encoder output, adapter output, and at an intermediate LLM layer. Representation drop-off at the adapter but deeper layers appear to resolve class confusions (e.g., wall vs. bed) and produce more spatially coherent predictions.
Figure 5: UMAP projections of patch-level hidden states across the MLLM stack. Each point represents one of 576 image patch tokens, colored by semantic class. At the adapter output, classes are interleaved, by layer 20, same-class patches form distinct clusters, illustrating the progressive emergence of semantic structure through the LLM layers.
Figure 6: Global attention knockout: layerwise segmentation comparison. Top row: unmodified baseline. Middle row: incorrect class blocked (e.g. sky). Bottom row: correct class blocked (e.g. ceiling). Blocking the incorrect class accelerates the resolution of misclassified patches across layers, while blocking the correct class impairs self-correction and leaves more errors in mid-to-late layers.
Figure 7: Context starvation under causal attention. Per-patch pixel accuracy for the first 50 tokens in the LLM layer of the MLLM. Accuracy Gap: Bidirectional minus causal accuracy. Early patches suffer severe context starvation under causal masking.
Figure 8: Causal vs. bidirectional attention: Layerwise probing on ADE20K across the MLLM Stack for different vision encoders. Bidirectional attention yields modest peak mIoU gains for CLIP and DINOv2, while SigLIP shows a larger effect.
Vision EncoderAttentionGQAMMBMME PMME CMMMUPOPESQA ITextVQAVizWiz
CLIP ViT-L/14Causal62.666.4148328435.386.868.746.956.0
Bidirectional62.765.5153828836.286.969.747.057.5
DINOv2 ViT-L/14Causal62.157.7130432434.687.266.114.051.4
Bidirectional60.555.3124732632.485.166.313.745.8
SigLIP SO400M/14Causal61.163.7141427534.784.470.450.258.2
Bidirectional62.166.9140229834.984.870.753.653.3
Table 1: VQA benchmark: causal vs. bidirectional LLaVA 1.5 (7B) across three vision encoders. All models use Vicuna-7B as the LLM backbone. Bidirectional attention preserves language understanding with segmentation gains that vary across encoders. DINOv2 shows broader regressions, consistent with its lack of text alignment.
Appendix figures & tables7 assets

Supplementary material from the paper’s appendix.

Appendix

Figure 9: Layerwise linear probing on Cityscapes (left) and PASCAL VOC (right). mIoU across the MLLM stack for CLIP, DINOv2, and SigLIP encoders paired with Vicuna-7B.
EncoderAttentionAdapterPeak LLMΔenc\Delta_{\text{enc}}Recovery
CLIP ViT-L/14@336Causal33.2240.74++ 6.36++ 7.52
Bidirectional32.6841.16++ 6.78++ 8.48
DINOv2 LargeCausal42.9044.50−- 1.08++ 1.60
Bidirectional40.4744.82−- 0.76++ 4.35
SigLIP SO400M/14Causal31.4435.39−- 3.31++ 3.95
Bidirectional32.3941.26++ 2.56++ 8.87
Appendix
Table 2: Causal vs. bidirectional attention: Layerwise probing on ADE20K. Adapter and Peak LLM report mIoU (%). Δenc\Delta_{\text{enc}} : Peak LLM mIoU minus vision encoder baseline (positive == surpasses encoder). Recovery: Peak LLM mIoU minus adapter mIoU.
 CausalBidirectionalΔ\Delta% Impr.
Patch 00.61950.7630++ 0.1435++ 23.2%
Patch 10.68510.7933++ 0.1082++ 15.8%
Patch 20.70690.7931++ 0.0862++ 12.2%
Appendix
Table 3: Context starvation under causal attention. Per-patch pixel accuracy for the first three tokens. Δ\Delta : Bidirectional minus causal accuracy. % Impr.: relative improvement over causal. Early patches suffer severe context starvation under causal masking.
Figure 10: Layerwise probing across different MLLMs. Δ\Delta mIoU relative to adapter output on ADE20K for LLaVA-1.5 (7B), LLaVA-OneVision (7B), and DeepSeek-VL (7B), each using its default vision encoder (CLIP for LLaVA-1.5, SigLIP for OneVision, SigLIP+SAM-B for DeepSeek-VL). Curves span different numbers of layers due to differences in LLM backbone depth. The adapter drop-off and LLM self-refinement pattern is consistent across architectures.
Figure 11: Aggregate attention knockout across three misclassified classes. Percentage of incorrectly predicted patches remaining across LLM layers, normalized to 100% at layer 0 and averaged over all images exhibiting the misclassification. Blocking the incorrect class (red) accelerates self-correction, blocking the correct class (blue) impairs it and can amplify errors beyond the initial count.
Figure 12: Qualitative attention knockout examples. Layerwise segmentation predictions under three conditions (rows: no knockout, knockout incorrect, knockout correct) for three class-confusion pairs. Top: sky vs. windowpane. Middle: ceiling vs. wall. Bottom: grass vs. road. Blocking the incorrect class produces cleaner segmentation maps at earlier layers, while blocking the correct class leaves errors unresolved or amplified.
Figure 13: Extended qualitative segmentation predictions across the MLLM stack. Seven ADE20K validation images showing, from left to right: input image, ground truth, linear probe prediction at the vision encoder output, adapter output, and at an intermediate LLM layer. The representation drop-off at the adapter is visible as increased noise and class confusion compared to the vision encoder output, while deeper LLM layers produce more spatially coherent predictions.
More in this topic

Multimodal Large Language Models

Multimodal large language models combine language models with image, video, audio, and other inputs for grounded understanding and generation.

See all 1,069 papers

Explore similar work

CardsList
Mar 19, 2026

Rethinking MLLM Itself as a Segmenter with a Single Segmentation Token

Recent segmentation methods leveraging Multi-modal Large Language Models (MLLMs) have shown reliable object-level segmentation and enhanced spatial perception. However, almost all previous methods predominantly rely on specialist mask decoders to interpret masks from generated segmentation-related embeddings and visual features, or incorporate multiple additional tokens to assist. This paper aims to investigate whether and how we can unlock segmentation from MLLM itSELF with 1 segmentation Embedding (SELF1E) while achieving competitive results, which eliminates the need for external decoders. To this end, our approach targets the fundamental limitation of resolution reduction in pixel-shuffled image features from MLLMs. First, we retain image features at their original uncompressed resolution, and refill them with residual features extracted from MLLM-processed compressed features, thereby improving feature precision. Subsequently, we integrate pixel-unshuffle operations on image features with and without LLM processing, respectively, to unleash the details of compressed features and amplify the residual features under uncompressed resolution, which further enhances the resolution of refilled features. Moreover, we redesign the attention mask with dual perception pathways, i.e., image-to-image and image-to-segmentation, enabling rich feature interaction between pixels and the segmentation token. Comprehensive experiments across multiple segmentation tasks validate that SELF1E achieves performance competitive with specialist mask decoder-based methods, demonstrating the feasibility of decoder-free segmentation in MLLMs. Project page: https://github.com/ANDYZAQ/SELF1E.

Anqi Zhang, Xiaokang Ji, Guangyu Gao +3

Beijing Institute of Technology · University of Birmingham · Beijing Jiaotong University +1

Image SegmentationSemantic SegmentationMultimodal Large Language Models
AskPDF
Feb 28, 2026

What Do Visual Tokens Really Encode? Uncovering Sparsity and Redundancy in Multimodal Large Language Models

Multimodal large language models (MLLMs) project visual tokens into the embedding space of language models, yet the internal structuring and processing of visual semantics remain poorly understood. In this work, we introduce a two-fold analytical framework featuring a novel probing tool, EmbedLens\textbf{EmbedLens}, to conduct a fine-grained analysis. We uncover a pronounced semantic sparsity at the input level: visual tokens consistently partition into sink, dead, and alive categories. Remarkably, only the alive tokens, comprising ≈60%\approx60\% of the total input, carry image-specific meaning. Furthermore, using a targeted patch-compression benchmark, we demonstrate that these alive tokens already encode rich, fine-grained cues (e.g., objects, colors, and OCR) prior to entering the LLM. Internal visual computations (such as visual attention and feed-forward networks) are redundant for most standard tasks. For the small subset of highly vision-centric tasks that actually benefit from internal processing, we reveal that alive tokens naturally align with intermediate LLM layers rather than the initial embedding space, indicating that shallow-layer processing is unnecessary and that direct mid-layer injection is both sufficient. Ultimately, our findings provide a unified mechanistic view of visual token processing, paving the way for more efficient and interpretable MLLM architectures through selective token pruning, minimized visual computation, and mid-layer injection. The code is released at: https://github.com/EIT-NLP/EmbedLens.

Yingqi Fan, Junlong Tong, Anhao Zhao +1

Institute of Digital Twin, Eastern Institute of Technology, Ningbo · Ningbo Key Laboratory of Spatial Intelligence and Digital Derivative · Shanghai Jiao Tong University +1

Multimodal Large Language ModelsMultimodal Model InterpretabilityVisual Token Pruning
AskPDF
Jul 7, 2026

Segmentation before Answering: Pixel Grounding for MLLM Visual Reasoning

Recent advancements in Multimodal Large Language Models (MLLMs) have evolved from static perception to interleaved visual-language reasoning, often referred to as ``thinking with images''. A basic operation in this reasoning process is to zoom in on regions of interest (often represented with bounding boxes) to acquire finer visual details. In this paper, we propose Segmentation before Answering (SegAnswer), which shifts the unit of zoom-in from the popular bounding box to pixel-level segmentation mask. By employing fine-grained masks to isolate the target area from cluttered environments, segmented visual input yields a more precise region of interest, effectively filtering out redundant background and interfering objects. Furthermore, the discrete patches of segmented visual input align more seamlessly with how MLLMs structure visual tokens via positional embeddings. In experiments, we evaluate SegAnswer across diverse benchmarks, including high-resolution perception, general perception, and hallucination. It achieves consistent improvements and also exhibits considerable performance on segmentation tasks, validating its capability for reliable pixel grounding.

Yake Wei, Yuan Wang, Fengyun Rao +2

Gaoling School of Artificial Intelligence, Renmin University of China · Laboratory of Research on Large Models and Intelligent Governance · Tsinghua University +1

Vision-Language GroundingVLM ReasoningVisually Grounded ReasoningVisual GroundingFine-Grained Visual Perception
AskPDF