Quantifying Modality Contributions via Disentangling Multimodal Representations

• URL: http://arxiv.org/abs/2511.19470v1
• Date: Sat, 22 Nov 2025 05:02:58 GMT
• Title: Quantifying Modality Contributions via Disentangling Multimodal Representations
• Authors: Padegal Amit, Omkar Mahesh Kashyap, Namitha Rayasam, Nidhi Shekhar, Surabhi Narayan,
• Abstract summary: Quantifying modality contributions in multimodal models remains a challenge, as existing approaches conflate the notion of contribution itself.<n>We propose a framework based on Partial Information Decomposition (PID) that quantifies modality contributions by decomposing predictive information in internal embeddings into unique, redundant, and synergistic components.<n>This provides a principled, representation-level view of multimodal behavior, offering clearer and more interpretable insights than outcome-based metrics.
• Score: 0.0
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Quantifying modality contributions in multimodal models remains a challenge, as existing approaches conflate the notion of contribution itself. Prior work relies on accuracy-based approaches, interpreting performance drops after removing a modality as indicative of its influence. However, such outcome-driven metrics fail to distinguish whether a modality is inherently informative or whether its value arises only through interaction with other modalities. This distinction is particularly important in cross-attention architectures, where modalities influence each other's representations. In this work, we propose a framework based on Partial Information Decomposition (PID) that quantifies modality contributions by decomposing predictive information in internal embeddings into unique, redundant, and synergistic components. To enable scalable, inference-only analysis, we develop an algorithm based on the Iterative Proportional Fitting Procedure (IPFP) that computes layer and dataset-level contributions without retraining. This provides a principled, representation-level view of multimodal behavior, offering clearer and more interpretable insights than outcome-based metrics.

Related papers

• Explaining multimodal LLMs via intra-modal token interactions [55.27436637894534]
  Multimodal Large Language Models (MLLMs) have achieved remarkable success across diverse vision-language tasks, yet their internal decision-making mechanisms remain insufficiently understood.<n>We propose enhancing interpretability by leveraging intra-modal interaction.
  arXiv  Detail & Related papers  (2025-09-26T14:39:13Z)
• Unified modality separation: A vision-language framework for unsupervised domain adaptation [60.8391821117794]
  Unsupervised domain adaptation (UDA) enables models trained on a labeled source domain to handle new unlabeled domains.<n>We propose a unified modality separation framework that accommodates both modality-specific and modality-invariant components.<n>Our methods achieve up to 9% performance gain with 9 times of computational efficiencies.
  arXiv  Detail & Related papers  (2025-08-07T02:51:10Z)
• Reducing Unimodal Bias in Multi-Modal Semantic Segmentation with Multi-Scale Functional Entropy Regularization [66.10528870853324]
  Fusing and balancing multi-modal inputs from novel sensors for dense prediction tasks is critically important.<n>One major limitation is the tendency of multi-modal frameworks to over-rely on easily learnable modalities.<n>We propose a plug-and-play regularization term based on functional entropy, which introduces no additional parameters.
  arXiv  Detail & Related papers  (2025-05-10T12:58:15Z)
• Multi-Modality Collaborative Learning for Sentiment Analysis [12.066757428026163]
  Multimodal sentiment analysis (MSA) identifies individuals' sentiment states in videos by integrating visual, audio, and text modalities.<n>Despite progress in existing methods, the inherent modality heterogeneity limits the effective capture of interactive sentiment features across modalities.<n>We introduce a Multi-Modality Collaborative Learning framework to facilitate cross-modal interactions and capture enhanced and complementary features from modality-common and modality-specific representations.
  arXiv  Detail & Related papers  (2025-01-21T12:06:21Z)
• Toward Robust Incomplete Multimodal Sentiment Analysis via Hierarchical Representation Learning [21.127950337002776]
  Multimodal Sentiment Analysis (MSA) is an important research area that aims to understand and recognize human sentiment through multiple modalities. We propose a Hierarchical Representation Learning Framework (HRLF) for the task under uncertain missing modalities. We show that HRLF significantly improves MSA performance under uncertain modality missing cases.
  arXiv  Detail & Related papers  (2024-11-05T04:04:41Z)
• Enhancing Unimodal Latent Representations in Multimodal VAEs through Iterative Amortized Inference [20.761803725098005]
  Multimodal variational autoencoders (VAEs) aim to capture shared latent representations by integrating information from different data modalities. A significant challenge is accurately inferring representations from any subset of modalities without training an impractical number of inference networks for all possible modality combinations. We introduce multimodal iterative amortized inference, an iterative refinement mechanism within the multimodal VAE framework.
  arXiv  Detail & Related papers  (2024-10-15T08:49:38Z)
• Mutual Information-based Representations Disentanglement for Unaligned Multimodal Language Sequences [25.73415065546444]
  Key challenge in unaligned multimodal language sequences is to integrate information from various modalities to obtain a refined multimodal joint representation. We propose a Mutual Information-based Representations Disentanglement (MIRD) method for unaligned multimodal language sequences.
  arXiv  Detail & Related papers  (2024-09-19T02:12:26Z)
• Out-of-Distribution Detection via Deep Multi-Comprehension Ensemble [11.542472900306745]
  Multi-Comprehension (MC) Ensemble is proposed as a strategy to augment the Out-of-Distribution (OOD) feature representation field. Our experimental results demonstrate the superior performance of the MC Ensemble strategy in OOD detection. This underscores the effectiveness of our proposed approach in enhancing the model's capability to detect instances outside its training distribution.
  arXiv  Detail & Related papers  (2024-03-24T18:43:04Z)
• Unified Multi-modal Unsupervised Representation Learning for Skeleton-based Action Understanding [62.70450216120704]
  Unsupervised pre-training has shown great success in skeleton-based action understanding. We propose a Unified Multimodal Unsupervised Representation Learning framework, called UmURL. UmURL exploits an efficient early-fusion strategy to jointly encode the multi-modal features in a single-stream manner.
  arXiv  Detail & Related papers  (2023-11-06T13:56:57Z)
• On Modality Bias Recognition and Reduction [70.69194431713825]
  We study the modality bias problem in the context of multi-modal classification. We propose a plug-and-play loss function method, whereby the feature space for each label is adaptively learned. Our method yields remarkable performance improvements compared with the baselines.
  arXiv  Detail & Related papers  (2022-02-25T13:47:09Z)
• Loss Bounds for Approximate Influence-Based Abstraction [81.13024471616417]
  Influence-based abstraction aims to gain leverage by modeling local subproblems together with the 'influence' that the rest of the system exerts on them. This paper investigates the performance of such approaches from a theoretical perspective. We show that neural networks trained with cross entropy are well suited to learn approximate influence representations.
  arXiv  Detail & Related papers  (2020-11-03T15:33:10Z)

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.