MIRROR: Aligning Semantic Relations from Language to Image via Gromov--Wasserstein

• URL: http://arxiv.org/abs/2606.29462v1
• Date: Sun, 28 Jun 2026 15:39:05 GMT
• Title: MIRROR: Aligning Semantic Relations from Language to Image via Gromov--Wasserstein
• Abstract summary: Multimodal Large Language Models (MLLMs) inherit rich relational priors from their language backbones, yet often fail when asked to apply these relationships in visual contexts.<n>We propose MIRROR, a geometric regularization framework that transfers relational priors from language to vision by exploiting the rich relational structure encoded in language representations.<n>We show that this formulation admits a unique closed-form solution that prescribes the ideal visual relational structure implied by language geometry and cross-modal coupling.
• Score: 20.422121064806188
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Multimodal Large Language Models (MLLMs) inherit rich relational priors from their language backbones, yet often fail when asked to apply these relationships in visual contexts. We trace this failure to a structural blind spot: projection-based alignment trains each visual token to carry the right semantics, but never asks whether the relationships between concepts survive the crossing from language to vision. To address this, we propose MIRROR (Mapping Inter-concept Relations from language to visual Representation via Optimal-transport-based Regularization), a geometric regularization framework that transfers relational priors from language to vision by exploiting the rich relational structure encoded in language representations. Specifically, we derive a surrogate loss from the proposed Semi-Inverse Gromov-Wasserstein (SI-GW) problem, an inverse geometric problem that aligns visual representations with language-derived relational priors. We show that this formulation admits a unique closed-form solution that prescribes the ideal visual relational structure implied by language geometry and cross-modal coupling. The structure of the formulation also enables efficient computation, making it applicable to long token sequences. Applying SI-GW inside decoder-only Transformers requires careful design. We introduce targeted strategies at the layer, head, and token levels to ensure stable extraction without additional parameters or inference cost. MIRROR improves relational consistency while preserving performance on general vision-language tasks.

Related papers

• Weave of Formal Thought [51.56484100374058]
  We introduce Weave of Formal Thought (WoFT), a paradigm uniting rigorous syntactic validation with learned structural representations.<n>We present a latent-variable fine-tuning method training the language model to interleave non-terminal grammar symbols directly into generation.<n>For Python, fine-tuning StarCoder2-3B with our RWS objective reduces per-token cross-entropy by 14.3% relative to a text-only SFT baseline.
  arXiv  Detail & Related papers  (2026-06-24T15:58:11Z)
• Improving Relative Representations with Learned Anchors and Whitened Inner Products [0.0]
  Independently trained neural models typically converge to incompatible latent representations.<n>Relative Representations (RR) address this by mapping absolute coordinates to a shared space defined by similarities to common anchor points.<n>Traditional implementations rely on randomly sampled anchors and cosine similarity.<n>We propose a robust framework for cross-model communication based on two improvements.
  arXiv  Detail & Related papers  (2026-05-28T21:43:53Z)
• Communication-Inspired Tokenization for Structured Image Representations [74.17163003465537]
  COMmunication inspired Tokenization (COMiT) is a framework for learning structured discrete visual token sequences.<n>Our experiments demonstrate that while semantic alignment provides grounding, attentive sequential tokenization is critical for inducing interpretable, object-centric token structure.
  arXiv  Detail & Related papers  (2026-02-24T09:53:50Z)
• Relation-R1: Progressively Cognitive Chain-of-Thought Guided Reinforcement Learning for Unified Relation Comprehension [31.952192907460713]
  Relation-R1 is the textitfirst unified relation comprehension framework.<n>It integrates cognitive chain-of-thought (CoT)-guided supervised fine-tuning (SFT) and group relative policy optimization ( GRPO)<n>Experiments on widely-used PSG and SWiG datasets demonstrate that Relation-R1 achieves state-of-the-art performance in both binary and textitN-ary relation understanding.
  arXiv  Detail & Related papers  (2025-04-20T14:50:49Z)
• PICASO: Permutation-Invariant Context Composition with State Space Models [98.91198288025117]
  State Space Models (SSMs) offer a promising solution by allowing a database of contexts to be mapped onto fixed-dimensional states.<n>We propose a simple mathematical relation derived from SSM dynamics to compose multiple states into one that efficiently approximates the effect of concatenating raw context tokens.<n>We evaluate our resulting method on WikiText and MSMARCO in both zero-shot and fine-tuned settings, and show that we can match the strongest performing baseline while enjoying on average 5.4x speedup.
  arXiv  Detail & Related papers  (2025-02-24T19:48:00Z)
• Cross-Modal Bidirectional Interaction Model for Referring Remote Sensing Image Segmentation [50.433911327489554]
  The goal of referring remote sensing image segmentation (RRSIS) is to generate a pixel-level mask of the target object identified by the referring expression.<n>To address the aforementioned challenges, a novel RRSIS framework is proposed, termed the cross-modal bidirectional interaction model (CroBIM)<n>To further forster the research of RRSIS, we also construct RISBench, a new large-scale benchmark dataset comprising 52,472 image-language-label triplets.
  arXiv  Detail & Related papers  (2024-10-11T08:28:04Z)
• Enhancing Language Representation with Constructional Information for Natural Language Understanding [5.945710973349298]
  We introduce construction grammar (CxG), which highlights the pairings of form and meaning. We adopt usage-based construction grammar as the basis of our work. A HyCxG framework is proposed to enhance language representation through a three-stage solution.
  arXiv  Detail & Related papers  (2023-06-05T12:15:12Z)
• Cross-modal Attention Congruence Regularization for Vision-Language Relation Alignment [105.70884254216973]
  We show that relation alignment can be enforced by encouraging the directed language attention from'mug' to 'grass' We prove that this notion of soft relation alignment is equivalent to enforcing congruence between vision and language attention. We apply our Cross-modal Attention Congruence Regularization (CACR) loss to UNITER and improve on the state-of-the-art approach to Winoground.
  arXiv  Detail & Related papers  (2022-12-20T18:53:14Z)
• SrTR: Self-reasoning Transformer with Visual-linguistic Knowledge for Scene Graph Generation [12.977857322594206]
  One-stage scene graph generation approaches infer the effective relation between entity pairs using sparse proposal sets and a few queries. A Self-reasoning Transformer with Visual-linguistic Knowledge (SrTR) is proposed to add flexible self-reasoning ability to the model. Inspired by the large-scale pre-training image-text foundation models, visual-linguistic prior knowledge is introduced.
  arXiv  Detail & Related papers  (2022-12-19T09:47:27Z)
• GATE: Graph Attention Transformer Encoder for Cross-lingual Relation and Event Extraction [107.8262586956778]
  We introduce graph convolutional networks (GCNs) with universal dependency parses to learn language-agnostic sentence representations. GCNs struggle to model words with long-range dependencies or are not directly connected in the dependency tree. We propose to utilize the self-attention mechanism to learn the dependencies between words with different syntactic distances.
  arXiv  Detail & Related papers  (2020-10-06T20:30:35Z)

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.