paper-with-me

홈 › Papers

The Indra Representation Hypothesis for Multimodal Alignment

2026-04-06 · Jianglin Lu, Hailing Wang, Kuo Yang, Yitian Zhang, Simon Jenni, Yun Fu arxiv

Recent studies have uncovered an interesting phenomenon: unimodal foundation models tend to learn convergent representations, regardless of differences in architecture, training objectives, or data modalities. However, these representations are essentially internal abstractions of samples that characterize samples independently, leading to limited expressiveness. In this paper, we propose The Indra Representation Hypothesis, inspired by the philosophical metaphor of Indra's Net. We argue that representations from unimodal foundation models are converging to implicitly reflect a shared relational structure underlying reality, akin to the relational ontology of Indra's Net. We formalize this hypothesis using the V-enriched Yoneda embedding from category theory, defining the Indra representation as a relational profile of each sample with respect to others. This formulation is shown to be unique, complete, and structure-preserving under a given cost function. We instantiate the Indra representation using angular distance and evaluate it in cross-model and cross-modal scenarios involving vision, language, and audio. Extensive experiments demonstrate that Indra representations consistently enhance robustness and alignment across architectures and modalities, providing a theoretically grounded and practical framework for training-free alignment of unimodal foundation models. Our code is available at https://github.com/Jianglin954/Indra.

📄 PDF Abstract BibTeX arXiv:2604.04496

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Mind the Heads: Topological Representation Alignment for Multimodal LLMs

2026-06-22 · Davide Caffagni, Alberto Compagnoni, Federico Melis, Sara Sarto 외 arxiv

Representation alignment has emerged as an effective approach to improve Multimodal Large Language Models (MLLMs) by regularizing their internal representations toward those of an external vision encoder. However, existi…

IndraEye: Infrared Electro-Optical UAV-based Perception Dataset for Robust Downstream Tasks

2024-10-28 · Manjunath D, Prajwal Gurunath, Sumanth Udupa, Aditya Gandhamal 외

Deep neural networks (DNNs) have shown exceptional performance when trained on well-illuminated images captured by Electro-Optical (EO) cameras, which provide rich texture details. However, in critical applications like …

Domain Adaptationobject-detectionObject DetectionSemantic Segmentation

Time Series, Vision, and Language: Exploring the Limits of Alignment in Contrastive Representation Spaces

2026-02-22 · Pratham Yashwante, Rose Yu arxiv

The Platonic Representation Hypothesis posits that learned representations from models trained on different modalities converge to a shared latent structure of the world. However, this hypothesis has largely been examine…

Contrastive Learning

Cross-Modal Prototype based Multimodal Federated Learning under Severely Missing Modality

2024-01-25 · Huy Q. Le, Chu Myaet Thwal, Yu Qiao, Ye Lin Tun 외

Multimodal federated learning (MFL) has emerged as a decentralized machine learning paradigm, allowing multiple clients with different modalities to collaborate on training a machine learning model across diverse data so…

cross-modal alignmentFederated Learning

Escaping Plato's Cave: JAM for Aligning Independently Trained Vision and Language Models

2025-07-01 · Lauren Hyoseo Yoon, Yisong Yue, Been Kim arxiv

Independently trained vision and language models inhabit disjoint representational spaces, shaped by their respective modalities, objectives, and architectures. The Platonic Representation Hypothesis (PRH) suggests these…