# Two Effects, One Trigger: On the Modality Gap, Object Bias, and Information Imbalance in Contrastive Vision-Language Models

arXiv: https://arxiv.org/abs/2404.07983 · HTML: https://mlsanity.com/paper/2404.07983/

## Key facts

- Cited by 57 papers in the map (data through 14 Oct 2026).
- Placed in Prompt learning and adaptation of CLIP, one of 16 directions in Multimodal LLMs / vision-language models.
- Builds on 49 of its 67 references that are in the map.
- First posted on arXiv 11 Apr 2024.

## Placed in

- Multimodal LLMs / vision-language models / [Prompt learning and adaptation of CLIP](https://mlsanity.com/mllm/prompt-learning-and-adaptation-of-clip/)

## Builds on (most cited first, 30 of 49)

- [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)
- [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [LAION-5B: An open large-scale dataset for training next generation image-text models](https://mlsanity.com/paper/2210.08402/) (2022)
- [Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision](https://mlsanity.com/paper/2102.05918/) (2021)
- [A ConvNet for the 2020s](https://mlsanity.com/paper/2201.03545/) (2022)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [UMAP: Uniform Manifold Approximation and Projection for Dimension Reduction](https://mlsanity.com/paper/1802.03426/) (2018)
- [Microsoft COCO Captions: Data Collection and Evaluation Server](https://mlsanity.com/paper/1504.00325/) (2015)
- [Understanding Contrastive Representation Learning through Alignment and Uniformity on the Hypersphere](https://mlsanity.com/paper/2005.10242/) (2020)
- [CoCa: Contrastive Captioners are Image-Text Foundation Models](https://mlsanity.com/paper/2205.01917/) (2022)
- [Conceptual 12M: Pushing Web-Scale Image-Text Pre-Training To Recognize Long-Tail Visual Concepts](https://mlsanity.com/paper/2102.08981/) (2021)
- [Reproducible scaling laws for contrastive language-image learning](https://mlsanity.com/paper/2212.07143/) (2022)
- [EVA-CLIP: Improved Training Techniques for CLIP at Scale](https://mlsanity.com/paper/2303.15389/) (2023)
- [PaLI: A Jointly-Scaled Multilingual Language-Image Model](https://mlsanity.com/paper/2209.06794/) (2022)
- [DataComp: In search of the next generation of multimodal datasets](https://mlsanity.com/paper/2304.14108/) (2023)
- [When and why vision-language models behave like bags-of-words, and what to do about it?](https://mlsanity.com/paper/2210.01936/) (2022)
- [Mind the Gap: Understanding the Modality Gap in Multi-modal Contrastive Representation Learning](https://mlsanity.com/paper/2203.02053/) (2022)
- [Visual Classification via Description from Large Language Models](https://mlsanity.com/paper/2210.07183/) (2022)
- [X-CLIP: End-to-End Multi-grained Contrastive Learning for Video-Text Retrieval](https://mlsanity.com/paper/2207.07285/) (2022)
- [Demystifying CLIP Data](https://mlsanity.com/paper/2309.16671/) (2023)
- [What does CLIP know about a red circle? Visual prompt engineering for VLMs](https://mlsanity.com/paper/2304.06712/) (2023)
- [Improving CLIP Training with Language Rewrites](https://mlsanity.com/paper/2305.20088/) (2023)
- [Partial success in closing the gap between human and machine vision](https://mlsanity.com/paper/2106.07411/) (2021)
- [CLIP2Scene: Towards Label-efficient 3D Scene Understanding by CLIP](https://mlsanity.com/paper/2301.04926/) (2023)
- [Evaluating CLIP: Towards Characterization of Broader Capabilities and Downstream Implications](https://mlsanity.com/paper/2108.02818/) (2021)
- [A Picture is Worth More Than 77 Text Tokens: Evaluating CLIP-Style Models on Dense Captions](https://mlsanity.com/paper/2312.08578/) (2023)

## Built on by (most cited first, 30 of 57)

- [jina-embeddings-v4: Universal Embeddings for Multimodal Multilingual Retrieval](https://mlsanity.com/paper/2506.18902/) (2025)
- [Same Task, Different Circuits: Disentangling Modality-Specific Mechanisms in VLMs](https://mlsanity.com/paper/2506.09047/) (2025)
- [When Vision Overrides Language: Evaluating and Mitigating Counterfactual Failures in VLAs](https://mlsanity.com/paper/2602.17659/) (2026)
- [Joint Vision-Language Social Bias Removal for CLIP](https://mlsanity.com/paper/2411.12785/) (2024)
- [Continual Learning for VLMs: A Survey and Taxonomy Beyond Forgetting](https://mlsanity.com/paper/2508.04227/) (2025)
- [VLM-Guard: Safeguarding Vision-Language Models via Fulfilling Safety Alignment Gap](https://mlsanity.com/paper/2502.10486/) (2025)
- [Cross-Modal Redundancy and the Geometry of Vision-Language Embeddings](https://mlsanity.com/paper/2602.06218/) (2026)
- [Reasoning Guided Embeddings: Leveraging MLLM Reasoning for Improved Multimodal Retrieval](https://mlsanity.com/paper/2511.16150/) (2025)
- [Enhancing Concept Localization in CLIP-based Concept Bottleneck Models](https://mlsanity.com/paper/2510.07115/) (2025)
- [Query Drift Compensation: Enabling Compatibility in Continual Learning of Retrieval Embedding Models](https://mlsanity.com/paper/2506.00037/) (2025)
- [Closing the Modality Gap Aligns Group-Wise Semantics](https://mlsanity.com/paper/2601.18525/) (2026)
- [Is the Modality Gap a Bug or a Feature? A Robustness Perspective](https://mlsanity.com/paper/2603.29080/) (2026)
- [Beyond Cross-Modal Alignment: Measuring and Leveraging Modality Gap in Vision-Language Models](https://mlsanity.com/paper/2502.14888/) (2025)
- [Multiplicity is an Inevitable and Inherent Challenge in Multimodal Learning](https://mlsanity.com/paper/2505.19614/) (2025)
- [Learning Shared Representations from Unpaired Data](https://mlsanity.com/paper/2505.21524/) (2025)
- [Common Data Properties Limit Object-Attribute Binding in CLIP](https://mlsanity.com/paper/2507.07985/) (2025)
- [D-TPT: Dimensional Entropy Maximization for Calibrating Test-Time Prompt Tuning in Vision-Language Models](https://mlsanity.com/paper/2510.09473/) (2025)
- [SpectralGCD: Spectral Concept Selection and Cross-modal Representation Learning for Generalized Category Discovery](https://mlsanity.com/paper/2602.17395/) (2026)
- [Pragma-VL: Towards a Pragmatic Arbitration of Safety and Helpfulness in MLLMs](https://mlsanity.com/paper/2603.13292/) (2026)
- [IsoCLIP: Decomposing CLIP Projectors for Efficient Intra-modal Alignment](https://mlsanity.com/paper/2603.19862/) (2026)
- [The Devil Is in the Details: Tackling Unimodal Spurious Correlations for Generalizable Multimodal Reward Models](https://mlsanity.com/paper/2503.03122/) (2025)
- [VisTIRA: Closing the Image-Text Modality Gap in Visual Math Reasoning via Structured Tool Integration](https://mlsanity.com/paper/2601.14440/) (2026)
- [Cross-Modal Prototype Alignment and Mixing for Training-Free Few-Shot Classification](https://mlsanity.com/paper/2603.24528/) (2026)
- [DETR-ViP: Detection Transformer with Robust Discriminative Visual Prompts](https://mlsanity.com/paper/2604.14684/) (2026)
- [Similarity Is Not Logic: Factored Inference for Dual-Encoder Vision-Language Models](https://mlsanity.com/paper/2607.23052/) (2026)
- [Fairness and Robustness of CLIP-Based Models for Chest X-rays](https://mlsanity.com/paper/2507.21291/) (2025)
- [Bridging the Modality Gap by Similarity Standardization with Pseudo-Positive Samples](https://mlsanity.com/paper/2511.22141/) (2025)
- [Closing the gap in multimodal medical representation alignment](https://mlsanity.com/paper/2602.20046/) (2026)
- [Dataset Diversity Metrics and Impact on Classification Models](https://mlsanity.com/paper/2603.15276/) (2026)
- [Bridging Domain Expertise and Generalization for Performance Estimation](https://mlsanity.com/paper/2606.06335/) (2026)

## Abstract (arXiv)

Contrastive vision-language models (VLMs), like CLIP, have gained popularity for their versatile applicability to various downstream tasks. Despite their successes in some tasks, like zero-shot object recognition, they perform surprisingly poor on other tasks, like attribute recognition. Previous work has attributed these challenges to the modality gap, a separation of image and text in the shared representation space, and to a bias towards objects over other factors, such as attributes. In this analysis paper, we investigate both phenomena thoroughly. We evaluated off-the-shelf VLMs and while the gap's influence on performance is typically overshadowed by other factors, we find indications that closing the gap indeed leads to improvements. Moreover, we find that, contrary to intuition, only few embedding dimensions drive the gap and that the embedding spaces are differently organized. To allow for a clean study of object bias, we introduce a definition and a corresponding measure of it. Equipped with this tool, we find that object bias does not lead to worse performance on other concepts, such as attributes per se. However, why do both phenomena, modality gap and object bias, emerge in the first place? To answer this fundamental question and uncover some of the inner workings of contrastive VLMs, we conducted experiments that allowed us to control the amount of shared information between the modalities. These experiments revealed that the driving factor behind both the modality gap and the object bias, is an information imbalance between images and captions, and unveiled an intriguing connection between the modality gap and entropy of the logits.