The Multi-Faceted Monosemanticity in Multimodal Representations

• URL: http://arxiv.org/abs/2502.14888v1
• Date: Sun, 16 Feb 2025 14:51:07 GMT
• Title: The Multi-Faceted Monosemanticity in Multimodal Representations
• Authors: Hanqi Yan, Xiangxiang Cui, Lu Yin, Paul Pu Liang, Yulan He, Yifei Wang,
• Abstract summary: We leverage recent advancements in feature monosemanticity to extract interpretable features from deep multimodal models.<n>Our findings reveal that this categorization aligns closely with human cognitive understandings of different modalities.<n>These results indicate that large-scale multimodal models, equipped with task-agnostic interpretability tools, offer valuable insights into key connections and distinctions between different modalities.
• Score: 42.64636740703632
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: In this paper, we leverage recent advancements in feature monosemanticity to extract interpretable features from deep multimodal models, offering a data-driven understanding of modality gaps. Specifically, we investigate CLIP (Contrastive Language-Image Pretraining), a prominent visual-language representation model trained on extensive image-text pairs. Building upon interpretability tools developed for single-modal models, we extend these methodologies to assess multi-modal interpretability of CLIP features. Additionally, we introduce the Modality Dominance Score (MDS) to attribute the interpretability of each feature to its respective modality. Next, we transform CLIP features into a more interpretable space, enabling us to categorize them into three distinct classes: vision features (single-modal), language features (single-modal), and visual-language features (cross-modal). Our findings reveal that this categorization aligns closely with human cognitive understandings of different modalities. We also demonstrate significant use cases of this modality-specific features including detecting gender bias, adversarial attack defense and text-to-image model editing. These results indicate that large-scale multimodal models, equipped with task-agnostic interpretability tools, offer valuable insights into key connections and distinctions between different modalities.

Related papers

• XR-VLM: Cross-Relationship Modeling with Multi-part Prompts and Visual Features for Fine-Grained Recognition [20.989787824067143]
  XR-VLM is a novel mechanism to discover subtle differences by modeling cross-relationships. We develop a multi-part prompt learning module to capture multi-perspective descriptions. Our method achieves significant improvements compared to current state-of-the-art approaches.
  arXiv  Detail & Related papers  (2025-03-10T08:58:05Z)
• A Survey on Mechanistic Interpretability for Multi-Modal Foundation Models [74.48084001058672]
  The rise of foundation models has transformed machine learning research. multimodal foundation models (MMFMs) pose unique interpretability challenges beyond unimodal frameworks. This survey explores two key aspects: (1) the adaptation of LLM interpretability methods to multimodal models and (2) understanding the mechanistic differences between unimodal language models and crossmodal systems.
  arXiv  Detail & Related papers  (2025-02-22T20:55:26Z)
• FSMR: A Feature Swapping Multi-modal Reasoning Approach with Joint Textual and Visual Clues [20.587249765287183]
  Feature Swapping Multi-modal Reasoning (FSMR) model is designed to enhance multi-modal reasoning through feature swapping. FSMR incorporates a multi-modal cross-attention mechanism, facilitating the joint modeling of textual and visual information. Experiments on the PMR dataset demonstrate FSMR's superiority over state-of-the-art baseline models.
  arXiv  Detail & Related papers  (2024-03-29T07:28:50Z)
• Multi-modal Auto-regressive Modeling via Visual Words [96.25078866446053]
  We propose the concept of visual tokens, which maps the visual features to probability distributions over Large Multi-modal Models' vocabulary. We further explore the distribution of visual features in the semantic space within LMM and the possibility of using text embeddings to represent visual information.
  arXiv  Detail & Related papers  (2024-03-12T14:58:52Z)
• Exploiting Modality-Specific Features For Multi-Modal Manipulation Detection And Grounding [54.49214267905562]
  We construct a transformer-based framework for multi-modal manipulation detection and grounding tasks. Our framework simultaneously explores modality-specific features while preserving the capability for multi-modal alignment. We propose an implicit manipulation query (IMQ) that adaptively aggregates global contextual cues within each modality.
  arXiv  Detail & Related papers  (2023-09-22T06:55:41Z)
• StableLLaVA: Enhanced Visual Instruction Tuning with Synthesized Image-Dialogue Data [129.92449761766025]
  We propose a novel data collection methodology that synchronously synthesizes images and dialogues for visual instruction tuning. This approach harnesses the power of generative models, marrying the abilities of ChatGPT and text-to-image generative models. Our research includes comprehensive experiments conducted on various datasets.
  arXiv  Detail & Related papers  (2023-08-20T12:43:52Z)
• UniDiff: Advancing Vision-Language Models with Generative and Discriminative Learning [86.91893533388628]
  This paper presents UniDiff, a unified multi-modal model that integrates image-text contrastive learning (ITC), text-conditioned image synthesis learning (IS), and reciprocal semantic consistency modeling (RSC) UniDiff demonstrates versatility in both multi-modal understanding and generative tasks.
  arXiv  Detail & Related papers  (2023-06-01T15:39:38Z)

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.