Multimodal Chain of Continuous Thought for Latent-Space Reasoning in Vision-Language Models

• URL: http://arxiv.org/abs/2508.12587v2
• Date: Tue, 23 Sep 2025 19:29:35 GMT
• Title: Multimodal Chain of Continuous Thought for Latent-Space Reasoning in Vision-Language Models
• Authors: Tan-Hanh Pham, Chris Ngo,
• Abstract summary: We propose the Multimodal Chain of Continuous Thought (MCOUT), which enables reasoning directly in a joint latent space rather than in natural language.<n>We show that MCOUT consistently improves multimodal reasoning, yielding up to 8.23% accuracy gains over strong baselines.<n>These findings highlight latent continuous reasoning as a promising direction for advancing LMMs beyond language-bound CoT.
• Score: 1.9950682531209158
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Many reasoning techniques for large multimodal models adapt language model approaches, such as Chain-of-Thought (CoT) prompting, which express reasoning as word sequences. While effective for text, these methods are suboptimal for multimodal contexts, struggling to align audio, visual, and textual information dynamically. To explore an alternative paradigm, we propose the Multimodal Chain of Continuous Thought (MCOUT), which enables reasoning directly in a joint latent space rather than in natural language. In MCOUT, the reasoning state is represented as a continuous hidden vector, iteratively refined and aligned with visual and textual embeddings, inspired by human reflective cognition. We develop two variants: MCOUT-Base, which reuses the language model`s last hidden state as the continuous thought for iterative reasoning, and MCOUT-Multi, which integrates multimodal latent attention to strengthen cross-modal alignment between visual and textual features. Experiments on benchmarks including MMMU, ScienceQA, and MMStar show that MCOUT consistently improves multimodal reasoning, yielding up to 8.23% accuracy gains over strong baselines and improving BLEU scores up to 8.27% across multiple-choice and open-ended tasks. These findings highlight latent continuous reasoning as a promising direction for advancing LMMs beyond language-bound CoT, offering a scalable framework for human-like reflective multimodal inference. Code is available at https://github.com/Hanhpt23/OmniMod.

Related papers

• Beyond Language Modeling: An Exploration of Multimodal Pretraining [125.34714978184638]
  We provide empirical clarity through controlled, from-scratch pretraining experiments.<n>We adopt the Transfusion framework, using next-token prediction for language and diffusion for vision.<n>We demonstrate that the MoE architecture harmonizes this scaling asymmetry by providing the high model capacity required by language.
  arXiv  Detail & Related papers  (2026-03-03T18:58:00Z)
• Reasoning Within the Mind: Dynamic Multimodal Interleaving in Latent Space [46.05748768260013]
  We propose a test-time Dynamic Multimodal Latent Reasoning framework.<n>It employs confidence-guided latent policy gradient optimization to latent think tokens for in-depth reasoning.<n> Experiments across seven multimodal reasoning benchmarks and various model architectures demonstrate that DMLR significantly improves reasoning and perception performance.
  arXiv  Detail & Related papers  (2025-12-14T10:07:45Z)
• CoCoVa: Chain of Continuous Vision-Language Thought for Latent Space Reasoning [22.835301879575002]
  CoCoVa is a novel framework for vision-language model that leverages continuous cross-modal reasoning for diverse vision-language tasks.<n>The core of CoCoVa is an iterative reasoning cycle, where a novel Latent Q-Former acts as a dynamic reasoning engine.<n>We train the model with a multi-task objective that combines contrastive learning and diffusion-based reconstruction.
  arXiv  Detail & Related papers  (2025-11-04T08:28:46Z)
• ThinkMorph: Emergent Properties in Multimodal Interleaved Chain-of-Thought Reasoning [76.95203056566191]
  Multimodal reasoning requires iterative coordination between language and vision, yet it remains unclear what constitutes a meaningful interleaved chain of thought.<n>We build ThinkMorph, a unified model fine-tuned on approximately 24K high-quality interleaved reasoning traces spanning tasks with varying visual engagement.<n>ThinkMorph learns to generate progressive text-image reasoning steps that concretely manipulate visual content while maintaining coherent verbal logic.
  arXiv  Detail & Related papers  (2025-10-30T17:51:38Z)
• MUCAR: Benchmarking Multilingual Cross-Modal Ambiguity Resolution for Multimodal Large Language Models [18.73221445082855]
  Multimodal Large Language Models (MLLMs) have demonstrated significant advances across numerous vision-language tasks.<n>We introduce MUCAR, a novel benchmark designed explicitly for evaluating multimodal ambiguity resolution across multilingual and cross-modal scenarios.
  arXiv  Detail & Related papers  (2025-06-20T14:57:41Z)
• Bridging the Dynamic Perception Gap: Training-Free Draft Chain-of-Thought for Dynamic Multimodal Spatial Reasoning [18.7712668000592]
  We present a novel maze navigation benchmark designed to evaluate dynamic spatial reasoning.<n>Experiments show that augmenting reasoning chains with dynamic visual drafts, overlaid on input images, significantly outperforms conventional approaches.<n>We propose D2R (Dynamic Draft-Augmented Reasoning), a training-free framework that seamlessly integrates textual CoT with corresponding visual drafts into MLLMs.
  arXiv  Detail & Related papers  (2025-05-22T12:14:23Z)
• Visual Thoughts: A Unified Perspective of Understanding Multimodal Chain-of-Thought [72.93910800095757]
  multimodal chain-of-thought (MCoT) improves performance and interpretability of large vision-language models (LVLMs)<n>We show that MCoT boosts LVLMs by incorporating visual thoughts, which convey image information to the reasoning process regardless of the MCoT format.<n>We also explore the internal nature of visual thoughts, finding that visual thoughts serve as intermediaries between the input image and reasoning to deeper transformer layers.
  arXiv  Detail & Related papers  (2025-05-21T13:29:58Z)
• CoMT: A Novel Benchmark for Chain of Multi-modal Thought on Large Vision-Language Models [60.08485416687596]
  Chain of Multi-modal Thought (CoMT) benchmark aims to mimic human-like reasoning that inherently integrates visual operation.<n>We evaluate various LVLMs and strategies on CoMT, revealing some key insights into the capabilities and limitations of the current approaches.
  arXiv  Detail & Related papers  (2024-12-17T14:10:16Z)
• MIO: A Foundation Model on Multimodal Tokens [75.45159462076937]
  We introduce MIO, a novel foundation model built on multimodal tokens.<n>MIO is capable of understanding and generating speech, text, images, and videos in an end-to-end, autoregressive manner.
  arXiv  Detail & Related papers  (2024-09-26T09:57:16Z)
• OmniBench: Towards The Future of Universal Omni-Language Models [63.16606414452612]
  We introduce OmniBench, a novel benchmark designed to evaluate models' ability to recognize, interpret, and reason across visual, acoustic, and textual inputs simultaneously.<n>Our evaluation reveals that open-source OLMs show significant limitations in instruction-following and reasoning in tri-modal contexts.<n>We advocate for developing more robust tri-modal integration techniques and training strategies to enhance OLM performance.
  arXiv  Detail & Related papers  (2024-09-23T17:59:05Z)
• Multimodal Chain-of-Thought Reasoning in Language Models [94.70184390935661]
  We propose Multimodal-CoT that incorporates language (text) and vision (images) modalities into a two-stage framework. Experimental results on ScienceQA and A-OKVQA benchmark datasets show the effectiveness of our proposed approach.
  arXiv  Detail & Related papers  (2023-02-02T07:51:19Z)

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.