Canvas-of-Thought: Grounding Reasoning via Mutable Structured States

• URL: http://arxiv.org/abs/2602.10494v1
• Date: Wed, 11 Feb 2026 03:54:58 GMT
• Title: Canvas-of-Thought: Grounding Reasoning via Mutable Structured States
• Abstract summary: Chain-of-Thought (CoT) prompting has significantly advanced the reasoning capabilities of Multimodal Large Language Models (MLLMs)<n> relying solely on linear text sequences remains a bottleneck for complex tasks.<n>We introduce textbf Canvas-of-Thought ( Canvas-CoT), a new paradigm for context-efficient multimodal reasoning.
• Score: 18.74176136703898
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While Chain-of-Thought (CoT) prompting has significantly advanced the reasoning capabilities of Multimodal Large Language Models (MLLMs), relying solely on linear text sequences remains a bottleneck for complex tasks. We observe that even when auxiliary visual elements are interleaved, they are often treated as static snapshots within a one-dimensional, unstructured reasoning chain. We argue that such approaches treat reasoning history as an immutable stream: correcting a local error necessitates either generating verbose downstream corrections or regenerating the entire context. This forces the model to implicitly maintain and track state updates, significantly increasing token consumption and cognitive load. This limitation is particularly acute in high-dimensional domains, such as geometry and SVG design, where the textual expression of CoT lacks explicit visual guidance, further constraining the model's reasoning precision. To bridge this gap, we introduce \textbf{Canvas-of-Thought (Canvas-CoT)}. By leveraging a HTML Canvas as an external reasoning substrate, Canvas-CoT empowers the model to perform atomic, DOM-based CRUD operations. This architecture enables in-place state revisions without disrupting the surrounding context, allowing the model to explicitly maintain the "ground truth". Furthermore, we integrate a rendering-based critique loop that serves as a hard constraint validator, providing explicit visual feedback to resolve complex tasks that are difficult to articulate through text alone. Extensive experiments on VCode, RBench-V, and MathVista demonstrate that Canvas-CoT significantly outperforms existing baselines, establishing a new paradigm for context-efficient multimodal reasoning.

Related papers

• CoVA-SFT: A Large-Scale Dataset for Chain of Visual Abstractions [73.85100620647452]
  Chain-of-thought (CoT) reasoning has dramatically improved large language models (LLMs)<n>CoVA-SFT is a highly structured corpus of 51.9K samples containing over 222K multimodal reasoning steps across 5 distinct layout families and 17 complex tasks.<n>By providing explicit rationale formulations, agentic renderings, and verification loops, CoVA-SFT teaches multimodal language models to interleave text and visual abstractions.
  arXiv  Detail & Related papers  (2026-08-29T00:03:13Z)
• ATLAS: Agentic or Latent Visual Reasoning? One Word is Enough for Both [55.182037225013836]
  ATLAS is a framework in which a single discrete 'word', termed as a functional token, serves both as an agentic operation and a latent visual reasoning unit.<n> ATLAS achieves superior performance on challenging benchmarks while maintaining clear interpretability.
  arXiv  Detail & Related papers  (2026-05-14T17:59:55Z)
• Decompose, Look, and Reason: Reinforced Latent Reasoning for VLMs [6.111899371682025]
  Vision-Language Models often struggle with complex visual reasoning due to the visual information loss in textual CoT.<n>We propose emph"Decompose, Look, and Reason" (DLR), a reinforced latent reasoning framework.<n>Experiments on vision-centric benchmarks show that DLR consistently outperforms strong baselines.
  arXiv  Detail & Related papers  (2026-04-08T18:52:27Z)
• Self-Corrected Image Generation with Explainable Latent Rewards [55.29175717238288]
  We propose xLARD, a self-correcting framework that guides generation through Explainable LAtent RewarDs.<n>xLARD introduces a lightweight corrector that refines latent representations based on structured feedback from model-generated references.<n> Experiments show that xLARD improves semantic alignment and visual fidelity while maintaining generative priors.
  arXiv  Detail & Related papers  (2026-03-26T02:59:35Z)
• From "What" to "How": Constrained Reasoning for Autoregressive Image Generation [26.716018030404665]
  CoR-Painter is a novel framework that pioneers a "How-to-What" paradigm by introducing Constrained Reasoning.<n>It first deduces "How to draw" by deriving a set of visual constraints from the input prompt.<n>These constraints steer the subsequent generation of a detailed description "What to draw", providing a structurally sound and coherent basis for accurate visual synthesis.
  arXiv  Detail & Related papers  (2026-03-03T08:03:18Z)
• Thinking with Images as Continuous Actions: Numerical Visual Chain-of-Thought [55.65577137924979]
  We propose a framework that enables MLLMs to reason over images using continuous numerical coordinates.<n> NV-CoT expands the MLLM action space from discrete vocabulary tokens to a continuous Euclidean space.<n>Experiments on three benchmarks demonstrate that NV-CoT significantly improves localization precision and final answer accuracy.
  arXiv  Detail & Related papers  (2026-02-27T12:04:07Z)
• On the Out-of-Distribution Generalization of Reasoning in Multimodal LLMs for Simple Visual Planning Tasks [56.98385132295952]
  We evaluate how well chain-of-thought approaches generalize on a simple planning task.<n>We find that reasoning traces which combine multiple text formats yield the best (and non-trivial) OOD generalization.<n> purely text-based models consistently outperform those utilizing image-based inputs.
  arXiv  Detail & Related papers  (2026-02-17T09:51:40Z)
• Render-of-Thought: Rendering Textual Chain-of-Thought as Images for Visual Latent Reasoning [23.364264811510598]
  Chain-of-Thought (CoT) prompting has achieved remarkable success in unlocking the reasoning capabilities of Large Language Models (LLMs)<n>We introduce Render-of-Thought (RoT), the first framework to reify the reasoning chain by rendering textual steps into images.<n>Our method achieves 3-4x token compression and substantial inference acceleration compared to explicit CoT.
  arXiv  Detail & Related papers  (2026-01-21T08:09:25Z)
• VisTIRA: Closing the Image-Text Modality Gap in Visual Math Reasoning via Structured Tool Integration [2.7403985180660784]
  Vision-language models (VLMs) lag behind text-only language models on mathematical reasoning when the same problems are presented as images rather than text.<n>We introduce VisTIRA, a tool-integrated reasoning framework that enables structured problem solving by decomposing a given math problem (as an image) into natural language rationales.<n>We show that tool-integrated supervision improves image-based reasoning, and OCR grounding can further narrow the gap for smaller models.
  arXiv  Detail & Related papers  (2026-01-20T19:54:49Z)
• Thinking with Deltas: Incentivizing Reinforcement Learning via Differential Visual Reasoning Policy [75.66913260900726]
  Reinforcement Learning with Verifiable Rewards has significantly advanced reasoning capabilities in Large Language Models.<n>Existing paradigms, driven by text-centric outcome rewards, encourage models to bypass visual perception.<n>We propose textbfThinking with Deltas, a framework driven by a textbfDifferential Visual Reasoning Policy.
  arXiv  Detail & Related papers  (2026-01-11T08:25:34Z)
• Figure It Out: Improve the Frontier of Reasoning with Executable Visual States [53.77871196174248]
  Complex reasoning problems often involve implicit spatial and geometric relationships that are not explicitly encoded in text.<n>We introduce FIGR, which integrates executable visual construction into multi-turn reasoning via end-to-end reinforcement learning.<n>Experiments on eight challenging mathematical benchmarks demonstrate that FIGR outperforms strong text-only chain-of-thought baselines.
  arXiv  Detail & Related papers  (2025-12-30T15:39:11Z)
• CRAFT: Continuous Reasoning and Agentic Feedback Tuning for Multimodal Text-to-Image Generation [0.0]
  CRAFT (Continuous Reasoning and Agentic Feedback Tuning) is a training-free, model-agnostic framework that brings structured reasoning paradigm to multimodal image generation.<n>It consistently improves compositional accuracy, text rendering, and preference-based evaluations.<n>These improvements incur only a negligible inference-time overhead, allowing smaller or cheaper models to approach the quality of substantially more expensive systems.
  arXiv  Detail & Related papers  (2025-12-23T13:44:41Z)
• CIR-CoT: Towards Interpretable Composed Image Retrieval via End-to-End Chain-of-Thought Reasoning [93.05917922306196]
  Composed Image Retrieval (CIR) aims to find a target image from a reference image and a modification text.<n>CIR-CoT is the first end-to-end retrieval-oriented MLLM designed to integrate explicit Chain-of-Thought (CoT) reasoning.
  arXiv  Detail & Related papers  (2025-10-09T09:41:45Z)
• Beyond the Linear Separability Ceiling: Aligning Representations in VLMs [2.561649173827544]
  We introduce a diagnostic framework centered on the Linear Separability Ceiling (LSC)<n>We uncover a pervasive "alignment gap", where most models fail to generatively outperform their own representations.<n>By augmenting standard next-token prediction with a contrastive objective, our fine-tuning method activates dormant reasoning pathways.
  arXiv  Detail & Related papers  (2025-07-10T09:23:32Z)
• ConText: Driving In-context Learning for Text Removal and Segmentation [59.6299939669307]
  This paper presents the first study on adapting the visual in-context learning paradigm to optical character recognition tasks.<n>We propose a task-chaining compositor in the form of image-removal-segmentation.<n>We also introduce context-aware aggregation, integrating the chained prompt pattern into the latent query representation.
  arXiv  Detail & Related papers  (2025-06-04T10:06:32Z)
• 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)
• Relation Rectification in Diffusion Model [64.84686527988809]
  We introduce a novel task termed Relation Rectification, aiming to refine the model to accurately represent a given relationship it initially fails to generate. We propose an innovative solution utilizing a Heterogeneous Graph Convolutional Network (HGCN) The lightweight HGCN adjusts the text embeddings generated by the text encoder, ensuring the accurate reflection of the textual relation in the embedding space.
  arXiv  Detail & Related papers  (2024-03-29T15:54:36Z)

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.