Understanding Counting Mechanisms in Large Language and Vision-Language Models

• URL: http://arxiv.org/abs/2511.17699v1
• Date: Fri, 21 Nov 2025 18:48:22 GMT
• Title: Understanding Counting Mechanisms in Large Language and Vision-Language Models
• Authors: Hosein Hasani, Amirmohammad Izadi, Fatemeh Askari, Mobin Bagherian, Sadegh Mohammadian, Mohammad Izadi, Mahdieh Soleymani Baghshah,
• Abstract summary: We study how large language models (LLMs) and large vision-language models (LVLMs) represent and compute numerical information in counting tasks.<n>Results show that individual tokens or visual features encode latent positional count information that can be extracted and transferred across contexts.<n>In LVLMs, numerical information also appears in visual embeddings, shifting between background and foreground regions depending on spatial composition.
• Score: 8.918147502104603
• License: http://creativecommons.org/licenses/by-nc-sa/4.0/
• Abstract: This paper examines how large language models (LLMs) and large vision-language models (LVLMs) represent and compute numerical information in counting tasks. We use controlled experiments with repeated textual and visual items and analyze model behavior through causal mediation and activation patching. To this end, we design a specialized tool, CountScope, for mechanistic interpretability of numerical content. Results show that individual tokens or visual features encode latent positional count information that can be extracted and transferred across contexts. Layerwise analyses reveal a progressive emergence of numerical representations, with lower layers encoding small counts and higher layers representing larger ones. We identify an internal counter mechanism that updates with each item, stored mainly in the final token or region and transferable between contexts. In LVLMs, numerical information also appears in visual embeddings, shifting between background and foreground regions depending on spatial composition. Models rely on structural cues such as separators in text, which act as shortcuts for tracking item counts and influence the accuracy of numerical predictions. Overall, counting emerges as a structured, layerwise process in LLMs and follows the same general pattern in LVLMs, shaped by the properties of the vision encoder.

Related papers

• Assessing the Visual Enumeration Abilities of Specialized Counting Architectures and Vision-Language Models [5.310444614342132]
  multimodal vision-language models (VLMs) may offer a flexible alternative for open-set object counting.<n>VLMs can approximately enumerate the number of items in a visual scene, matching or even surpassing the performance of specialized computer vision architectures.<n>None of the models can reliably count the number of objects in complex visual scenes.
  arXiv  Detail & Related papers  (2025-12-17T09:56:25Z)
• Improving Large Vision-Language Models' Understanding for Field Data [62.917026891829025]
  We introduce FieldLVLM, a framework designed to improve large vision-language models' understanding of field data.<n>FieldLVLM consists of two main components: a field-aware language generation strategy and a data-compressed multimodal model tuning.<n> Experimental results on newly proposed benchmark datasets demonstrate that FieldLVLM significantly outperforms existing methods in tasks involving scientific field data.
  arXiv  Detail & Related papers  (2025-07-24T11:28:53Z)
• FOCUS: Unified Vision-Language Modeling for Interactive Editing Driven by Referential Segmentation [55.01077993490845]
  Recent Large Vision Language Models (LVLMs) demonstrate promising capabilities in unifying visual understanding and generative modeling.<n>We introduce FOCUS, a unified LVLM that integrates segmentation-aware perception and controllable object-centric generation within an end-to-end framework.
  arXiv  Detail & Related papers  (2025-06-20T07:46:40Z)
• AlignVLM: Bridging Vision and Language Latent Spaces for Multimodal Document Understanding [79.43306110124875]
  AlignVLM is a vision-text alignment method that maps visual features to a weighted average of text embeddings.<n>Our experiments show that AlignVLM achieves state-of-the-art performance compared to prior alignment methods.
  arXiv  Detail & Related papers  (2025-02-03T13:34:51Z)
• Seeing Syntax: Uncovering Syntactic Learning Limitations in Vision-Language Models [18.87130615326443]
  Vision-language models (VLMs) serve as foundation models for image captioning and text-to-image generation.<n>Recent studies have highlighted limitations in VLM text encoders, particularly in areas like compositionality and semantic understanding.
  arXiv  Detail & Related papers  (2024-12-11T05:37:04Z)
• Cross-modal Information Flow in Multimodal Large Language Models [14.853197288189579]
  We investigate the information flow between different modalities -- language and vision -- in large language models (MLLMs)<n>We find that there are two distinct stages in the process of integration of the two modalities.<n>Our findings provide a new and comprehensive perspective on the spatial and functional aspects of image and language processing in the MLLMs.
  arXiv  Detail & Related papers  (2024-11-27T18:59:26Z)
• Towards Interpreting Visual Information Processing in Vision-Language Models [24.51408101801313]
  Vision-Language Models (VLMs) are powerful tools for processing and understanding text and images.<n>We study the processing of visual tokens in the language model component of LLaVA, a prominent VLM.
  arXiv  Detail & Related papers  (2024-10-09T17:55:02Z)
• Vector-ICL: In-context Learning with Continuous Vector Representations [75.96920867382859]
  Large language models (LLMs) have shown remarkable in-context learning capabilities on textual data.<n>We explore whether these capabilities can be extended to continuous vectors from diverse domains, obtained from black-box pretrained encoders.<n>In particular, we find that pretraining projectors with general language modeling objectives enables Vector-ICL.
  arXiv  Detail & Related papers  (2024-10-08T02:25:38Z)
• Large Language Models Understand Layout [6.732578061359833]
  Large language models (LLMs) demonstrate extraordinary abilities in a wide range of natural language processing (NLP) tasks. We show that, beyond text understanding capability, LLMs are capable of processing text layouts denoted by spatial markers. We show that layout understanding ability is beneficial for building efficient visual question-answering (VQA) systems.
  arXiv  Detail & Related papers  (2024-07-08T09:03:12Z)
• CODIS: Benchmarking Context-Dependent Visual Comprehension for Multimodal Large Language Models [58.95889895912716]
  We introduce a new benchmark, named as CODIS, designed to assess the ability of models to use context provided in free-form text to enhance visual comprehension. Our findings indicate that MLLMs consistently fall short of human performance on this benchmark. This underscores the pressing need to enhance the ability of MLLMs to comprehend visuals in a context-dependent manner.
  arXiv  Detail & Related papers  (2024-02-21T08:21:12Z)
• RefSAM: Efficiently Adapting Segmenting Anything Model for Referring Video Object Segmentation [53.4319652364256]
  This paper presents the RefSAM model, which explores the potential of SAM for referring video object segmentation. Our proposed approach adapts the original SAM model to enhance cross-modality learning by employing a lightweight Cross-RValModal. We employ a parameter-efficient tuning strategy to align and fuse the language and vision features effectively.
  arXiv  Detail & Related papers  (2023-07-03T13:21:58Z)
• CLIP-Count: Towards Text-Guided Zero-Shot Object Counting [32.07271723717184]
  We propose CLIP-Count, the first end-to-end pipeline that estimates density maps for open-vocabulary objects with text guidance in a zero-shot manner. To align the text embedding with dense visual features, we introduce a patch-text contrastive loss that guides the model to learn informative patch-level visual representations for dense prediction. Our method effectively generates high-quality density maps for objects-of-interest.
  arXiv  Detail & Related papers  (2023-05-12T08:19:39Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.