The Geometry of Representational Failures in Vision Language Models

• URL: http://arxiv.org/abs/2602.07025v1
• Date: Mon, 02 Feb 2026 12:20:04 GMT
• Title: The Geometry of Representational Failures in Vision Language Models
• Authors: Daniele Savietto, Declan Campbell, André Panisson, Marco Nurisso, Giovanni Petri, Jonathan D. Cohen, Alan Perotti,
• Abstract summary: Vision-Language Models (VLMs) exhibit puzzling failures in multi-object visual tasks.<n>These errors mirror human cognitive constraints, such as the "Binding Problem"<n>We propose a mechanistic insight by analyzing the representational geometry of open-weight VLMs.
• Score: 5.7337123720860435
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Vision-Language Models (VLMs) exhibit puzzling failures in multi-object visual tasks, such as hallucinating non-existent elements or failing to identify the most similar objects among distractions. While these errors mirror human cognitive constraints, such as the "Binding Problem", the internal mechanisms driving them in artificial systems remain poorly understood. Here, we propose a mechanistic insight by analyzing the representational geometry of open-weight VLMs (Qwen, InternVL, Gemma), comparing methodologies to distill "concept vectors" - latent directions encoding visual concepts. We validate our concept vectors via steering interventions that reliably manipulate model behavior in both simplified and naturalistic vision tasks (e.g., forcing the model to perceive a red flower as blue). We observe that the geometric overlap between these vectors strongly correlates with specific error patterns, offering a grounded quantitative framework to understand how internal representations shape model behavior and drive visual failures.

Related papers

• From Perception to Action: An Interactive Benchmark for Vision Reasoning [51.11355591375073]
  Causal Hierarchy of Actions and Interactions (CHAIN) benchmark designed to evaluate whether models can understand, plan, and execute structured action sequences grounded in physical constraints.<n> CHAIN shifts evaluation from passive perception to active problem solving, spanning tasks such as interlocking mechanical puzzles and 3D stacking and packing.<n>Our results show that top-performing models still struggle to internalize physical structure and causal constraints, often failing to produce reliable long-horizon plans and cannot robustly translate perceived structure into effective actions.
  arXiv  Detail & Related papers  (2026-02-24T15:33:02Z)
• MentisOculi: Revealing the Limits of Reasoning with Mental Imagery [63.285794947638614]
  We develop MentisOculi, a suite of multi-step reasoning problems amenable to visual solution.<n> evaluating visual strategies ranging from latent tokens to explicit generated imagery, we find they generally fail to improve performance.<n>Our findings suggest that despite their inherent appeal, visual thoughts do not yet benefit model reasoning.
  arXiv  Detail & Related papers  (2026-02-02T18:49:06Z)
• LVLM-Aided Alignment of Task-Specific Vision Models [49.96265491629163]
  Small task-specific vision models are crucial in high-stakes domains.<n>We introduce a novel and efficient method for aligning small task-specific vision models with human domain knowledge.<n>Our method demonstrates substantial improvement in aligning model behavior with human specifications.
  arXiv  Detail & Related papers  (2025-12-26T11:11:25Z)
• Sparse Autoencoders for Scientifically Rigorous Interpretation of Vision Models [27.806966289284528]
  We present a unified framework using sparse autoencoders (SAEs) to discover human-interpretable visual features.<n>We show that SAEs can reliably identify and manipulate interpretable visual features without model re-training.
  arXiv  Detail & Related papers  (2025-02-10T18:32:41Z)
• A Cognitive Paradigm Approach to Probe the Perception-Reasoning Interface in VLMs [3.2228025627337864]
  This paper introduces a structured evaluation framework to dissect the perception-reasoning interface in Vision-Language Models (VLMs)<n>We propose three distinct evaluation paradigms, mirroring human problem-solving strategies.<n>Applying this framework, we demonstrate that CA, leveraging powerful language models for reasoning over rich, independently generated descriptions, achieves new state-of-the-art (SOTA) performance.
  arXiv  Detail & Related papers  (2025-01-23T12:42:42Z)
• Understanding the Limits of Vision Language Models Through the Lens of the Binding Problem [37.27516441519387]
  We show that state-of-the-art vision language models exhibit surprising failures on basic multi-object reasoning tasks that humans perform with near perfect accuracy.<n>We find that many of the puzzling failures of state-of-the-art VLMs can be explained as arising due to the binding problem, and that these failure modes are strikingly similar to the limitations exhibited by rapid, feedforward processing in the human brain.
  arXiv  Detail & Related papers  (2024-10-31T22:24:47Z)
• Automatic Discovery of Visual Circuits [66.99553804855931]
  We explore scalable methods for extracting the subgraph of a vision model's computational graph that underlies recognition of a specific visual concept. We find that our approach extracts circuits that causally affect model output, and that editing these circuits can defend large pretrained models from adversarial attacks.
  arXiv  Detail & Related papers  (2024-04-22T17:00:57Z)
• CogCoM: A Visual Language Model with Chain-of-Manipulations Reasoning [61.21923643289266]
  Chain of Manipulations is a mechanism that enables Vision-Language Models to solve problems step-by-step with evidence.<n>After training, models can solve various visual problems by eliciting intrinsic manipulations (e.g., grounding, zoom in) actively without involving external tools.<n>Our trained model, textbfCogCoM, achieves state-of-the-art performance across 9 benchmarks from 4 categories.
  arXiv  Detail & Related papers  (2024-02-06T18:43:48Z)
• Internal Representations of Vision Models Through the Lens of Frames on Data Manifolds [8.67467876089153]
  We present a new approach to studying such representations inspired by the idea of a frame on the tangent bundle of a manifold. Our construction, which we call a neural frame, is formed by assembling a set of vectors representing specific types of perturbations of a data point. Using neural frames, we make observations about the way that models process, layer-by-layer, specific modes of variation within a small neighborhood of a datapoint.
  arXiv  Detail & Related papers  (2022-11-19T01:48:19Z)
• Visual Perturbation-aware Collaborative Learning for Overcoming the Language Prior Problem [60.0878532426877]
  We propose a novel collaborative learning scheme from the viewpoint of visual perturbation calibration. Specifically, we devise a visual controller to construct two sorts of curated images with different perturbation extents. The experimental results on two diagnostic VQA-CP benchmark datasets evidently demonstrate its effectiveness.
  arXiv  Detail & Related papers  (2022-07-24T23:50:52Z)
• Interpretable Visual Reasoning via Induced Symbolic Space [75.95241948390472]
  We study the problem of concept induction in visual reasoning, i.e., identifying concepts and their hierarchical relationships from question-answer pairs associated with images. We first design a new framework named object-centric compositional attention model (OCCAM) to perform the visual reasoning task with object-level visual features. We then come up with a method to induce concepts of objects and relations using clues from the attention patterns between objects' visual features and question words.
  arXiv  Detail & Related papers  (2020-11-23T18:21:49Z)

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.