Performance Gap in Entity Knowledge Extraction Across Modalities in Vision Language Models
- URL: http://arxiv.org/abs/2412.14133v1
- Date: Wed, 18 Dec 2024 18:22:30 GMT
- Title: Performance Gap in Entity Knowledge Extraction Across Modalities in Vision Language Models
- Authors: Ido Cohen, Daniela Gottesman, Mor Geva, Raja Giryes,
- Abstract summary: Vision-language models (VLMs) excel at extracting and reasoning about information from images.
This work investigates the disparity in model performance when answering factual questions about an entity described in text versus depicted in an image.
- Score: 36.18155629835474
- License:
- Abstract: Vision-language models (VLMs) excel at extracting and reasoning about information from images. Yet, their capacity to leverage internal knowledge about specific entities remains underexplored. This work investigates the disparity in model performance when answering factual questions about an entity described in text versus depicted in an image. Our results reveal a significant accuracy drop --averaging 19%-- when the entity is presented visually instead of textually. We hypothesize that this decline arises from limitations in how information flows from image tokens to query tokens. We use mechanistic interpretability tools to reveal that, although image tokens are preprocessed by the vision encoder, meaningful information flow from these tokens occurs only in the much deeper layers. Furthermore, critical image processing happens in the language model's middle layers, allowing few layers for consecutive reasoning, highlighting a potential inefficiency in how the model utilizes its layers for reasoning. These insights shed light on the internal mechanics of VLMs and offer pathways for enhancing their reasoning capabilities.
Related papers
- DRIVINGVQA: Analyzing Visual Chain-of-Thought Reasoning of Vision Language Models in Real-World Scenarios with Driving Theory Tests [69.00444996464662]
We present DrivingVQA, a new benchmark derived from driving theory tests to evaluate visual chain-of-thought reasoning in complex real-world scenarios.
Our experiments reveal that open-source and proprietary LVLMs struggle with visual chain-of-thought reasoning under zero-shot settings.
We investigate training strategies that leverage relevant entities to improve visual reasoning.
arXiv Detail & Related papers (2025-01-08T18:31:16Z) - What's in the Image? A Deep-Dive into the Vision of Vision Language Models [20.669971132114195]
Vision-Language Models (VLMs) have recently demonstrated remarkable capabilities in comprehending complex visual content.
In this paper, we conduct a thorough empirical analysis, focusing on attention modules across layers.
We reveal several key insights about how these models process visual data.
arXiv Detail & Related papers (2024-11-26T14:59:06Z) - 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.
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) - VALE: A Multimodal Visual and Language Explanation Framework for Image Classifiers using eXplainable AI and Language Models [0.0]
We propose a novel framework named VALE Visual and Language Explanation.
VALE integrates explainable AI techniques with advanced language models to provide comprehensive explanations.
In this paper, we conduct a pilot study of the VALE framework for image classification tasks.
arXiv Detail & Related papers (2024-08-23T03:02:11Z) - Enhancing Large Vision Language Models with Self-Training on Image Comprehension [131.14381425260706]
We introduce Self-Training on Image (STIC), which emphasizes a self-training approach specifically for image comprehension.
First, the model self-constructs a preference for image descriptions using unlabeled images.
To further self-improve reasoning on the extracted visual information, we let the model reuse a small portion of existing instruction-tuning data.
arXiv Detail & Related papers (2024-05-30T05:53:49Z) - InDL: A New Dataset and Benchmark for In-Diagram Logic Interpretation
based on Visual Illusion [1.7980584146314789]
This paper introduces a novel approach to evaluating deep learning models' capacity for in-diagram logic interpretation.
We establish a unique dataset, InDL, designed to rigorously test and benchmark these models.
We utilize six classic geometric optical illusions to create a comparative framework between human and machine visual perception.
arXiv Detail & Related papers (2023-05-28T13:01:32Z) - Perceptual Grouping in Contrastive Vision-Language Models [59.1542019031645]
We show how vision-language models are able to understand where objects reside within an image and group together visually related parts of the imagery.
We propose a minimal set of modifications that results in models that uniquely learn both semantic and spatial information.
arXiv Detail & Related papers (2022-10-18T17:01:35Z) - Exploring CLIP for Assessing the Look and Feel of Images [87.97623543523858]
We introduce Contrastive Language-Image Pre-training (CLIP) models for assessing both the quality perception (look) and abstract perception (feel) of images in a zero-shot manner.
Our results show that CLIP captures meaningful priors that generalize well to different perceptual assessments.
arXiv Detail & Related papers (2022-07-25T17:58:16Z) - Generative Counterfactuals for Neural Networks via Attribute-Informed
Perturbation [51.29486247405601]
We design a framework to generate counterfactuals for raw data instances with the proposed Attribute-Informed Perturbation (AIP)
By utilizing generative models conditioned with different attributes, counterfactuals with desired labels can be obtained effectively and efficiently.
Experimental results on real-world texts and images demonstrate the effectiveness, sample quality as well as efficiency of our designed framework.
arXiv Detail & Related papers (2021-01-18T08:37:13Z)
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.