mlsanityData to 14 Oct 2026

OneLatent: Single-Token Compression for Visual Latent Reasoning

arXiv 2602.13738 ↗First posted 14 Feb 2026cs.AIcited by 5

On the map: LLM reasoning → Multimodal reasoning

also in 1 more
Abstract (arXiv)

Chain-of-thought (CoT) prompting improves reasoning but often increases inference cost by one to two orders of magnitude. To address these challenges, we present OneLatent, a framework that compresses intermediate reasoning into a single latent token via supervision from rendered CoT images and DeepSeek-OCR hidden states. By rendering textual steps into images, we obtain a deterministic supervision signal that can be inspected and audited without requiring the model to output verbose textual rationales. Across benchmarks, OneLatent reduces average output length by 11× with only a 2.21% average accuracy drop relative to textual CoT, while improving output token contribution (OTC) by 6.8×. On long-chain logical reasoning, OneLatent reaches 99.80% on ProntoQA and 97.80% on ProsQA with one latent token, with compression up to 87.4×, supporting compression-constrained generalization.

Key facts. Cited by 5 papers in the map (data through 14 Oct 2026). Placed in Multimodal reasoning, one of 17 directions in LLM reasoning. Builds on 29 of its 30 references that are in the map. First posted on arXiv 14 Feb 2026.

Builds on

29 of 30 references.

References · 29

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,744
  2. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  3. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  4. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  5. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,276
  6. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
  7. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,211
  8. 2022Least-to-Most Prompting Enables Complex Reasoning in Large Language Modelscited by 1,407
  9. 2023Language Models Don't Always Say What They Think: Unfaithful Explanations in Chain-of-Thought Promptingcited by 1,122
  10. 2023Math-Shepherd: Verify and Reinforce LLMs Step-by-step without Human Annotationscited by 928
  11. 2023Reasoning with Language Model is Planning with World Modelcited by 861
  12. 2024Training Large Language Models to Reason in a Continuous Latent Spacecited by 797
17 more
  1. 2022Decomposed Prompting: A Modular Approach for Solving Complex Taskscited by 558
  2. 2023On the Planning Abilities of Large Language Models : A Critical Investigationcited by 312
  3. 2024Quiet-STaR: Language Models Can Teach Themselves to Think Before Speakingcited by 301
  4. 2023Think before you speak: Training Language Models With Pause Tokenscited by 264
  5. 2024To CoT or not to CoT? Chain-of-thought helps mainly on math and symbolic reasoningcited by 263
  6. 2024From Explicit CoT to Implicit CoT: Learning to Internalize CoT Step by Stepcited by 235
  7. 2022Towards Understanding Chain-of-Thought Prompting: An Empirical Study of What Matterscited by 222
  8. 2023Self-Evaluation Guided Beam Search for Reasoningcited by 196
  9. 2024Let's Think Dot by Dot: Hidden Computation in Transformer Language Modelscited by 191
  10. 2023Implicit Chain of Thought Reasoning via Knowledge Distillationcited by 176
  11. 2024From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Modelscited by 137
  12. 2022Text and Patterns: For Effective Chain of Thought, It Takes Two to Tangocited by 128
  13. 2024LLM Reasoners: New Evaluation, Library, and Analysis of Step-by-Step Reasoning with Large Language Modelscited by 70
  14. 2025SIM-CoT: Supervised Implicit Chain-of-Thoughtcited by 70
  15. 2023Guiding Language Model Reasoning with Planning Tokenscited by 58
  16. 2025Glyph: Scaling Context Windows via Visual-Text Compressioncited by 49
  17. 2019Minimum Description Length Revisitedcited by 37

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 5.

5 papers.

  1. 2026The Latent Space: Foundation, Evolution, Mechanism, Ability, and Outlookcited by 43
  2. 2026UniVLR: Unifying Text and Vision in Visual Latent Reasoning for Multimodal LLMscited by 8
  3. 2026Selective Latent Thinking: Adaptive Compression of LLM Reasoning Chainscited by 2
  4. 2026Why Struggle with Continuous Latents? Interpretable Discrete Latent Reasoning via Rendered Compressioncited by 2
  5. 2026DeepLatent: Think with Images via Parallel Latent Visual Reasoningcited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2026Render-of-Thought: Rendering Textual Chain-of-Thought as Images for Visual Latent Reasoning
  2. 2026Rethinking Latent Visual Reasoning: Grounding Latent Reasoning in Visual Evidence
  3. 2026VTC-R1: Vision-Text Compression for Efficient Long-Context Reasoning
  4. 2026CoLT: Teaching Multi-Modal Models to Think with Chain of Latent Thoughts
  5. 2026ReGuLaR: Variational Latent Reasoning Guided by Rendered Chain-of-Thought
  6. 2026SYNLAT: Syntax-Aligned Text-Latent Compression for Chain-of-Thought Reasoning
  7. 2026Self-Consistent Latent Reasoning: Long Latent Sequence Reasoning for Vision-Language Model
  8. 2026VIG: Visual Information Gain as a Reward Signal for Multimodal Chain-of-Thought Compression