mlsanityData to 14 Oct 2026

DCGC: Draft-Conditioned Global Correction for Complex Reasoning with Masked Diffusion Models

arXiv 2608.25428 ↗First posted 26 Aug 2026cs.CL cs.AIcited by 0

On the map: LLM reasoning → Diffusion reasoning and test-time scaling

also in 1 more
Abstract (arXiv)

Correcting flawed reasoning traces remains a significant challenge for Large Language Models (LLMs), whose autoregressive generation can propagate early mistakes into subsequent reasoning. We introduce DCGC, a Masked Diffusion Model (MDM) framework for global correction that uses an imperfect solution draft from an upstream solver as auxiliary context. DCGC combines task-specific Supervised Fine-Tuning (SFT) with a novel inference-time mechanism called Dynamic Dual-CFG. This mechanism separates problem-only and joint problem-draft branches and scales the draft-conditioned residual using a relative confidence gap. Across math, code, and knowledge reasoning benchmarks, DCGC outperforms standard sampling and simpler CFG variants, with additional results suggesting transfer to different diffusion backbones. In test-time setting where ground-truth failure labels are unavailable, DCGC improves full test set accuracy by correcting low-consensus upstream outputs, highlighting its utility as a verifier-free global correction module for difficult reasoning instances.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Diffusion reasoning and test-time scaling, one of 17 directions in LLM reasoning. Builds on 44 of its 46 references that are in the map. First posted on arXiv 26 Aug 2026.

Builds on

44 of 46 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025Mercury: Ultra-Fast Language Models Based on Diffusion17 Jun
  2. 2024Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning18 Oct

Other references · 42

  1. 2024The Llama 3 Herd of Modelscited by 14,715
  2. 2021LoRA: Low-Rank Adaptation of Large Language Modelscited by 14,380
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  4. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  5. 2021Evaluating Large Language Models Trained on Codecited by 8,320
  6. 2020Measuring Massive Multitask Language Understandingcited by 7,963
  7. 2022Classifier-Free Diffusion Guidancecited by 5,847
  8. 2022Self-Consistency Improves Chain of Thought Reasoning in Language Modelscited by 5,708
  9. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  10. 2017On Calibration of Modern Neural Networkscited by 4,960
  11. 2023Reflexion: Language Agents with Verbal Reinforcement Learningcited by 4,276
  12. 2021Program Synthesis with Large Language Modelscited by 3,637
30 more
  1. 2023Self-Refine: Iterative Refinement with Self-Feedbackcited by 3,442
  2. 2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,211
  3. 2023Mistral 7Bcited by 3,164
  4. 2022InstructPix2Pix: Learning to Follow Image Editing Instructionscited by 2,611
  5. 2022Language Models (Mostly) Know What They Knowcited by 1,856
  6. 2023Improving Factuality and Reasoning in Language Models through Multiagent Debatecited by 1,784
  7. 2021Structured Denoising Diffusion Models in Discrete State-Spacescited by 1,719
  8. 2022Diffusion-LM Improves Controllable Text Generationcited by 1,027
  9. 2023Large Language Models Cannot Self-Correct Reasoning Yetcited by 1,017
  10. 2023Math-Shepherd: Verify and Reinforce LLMs Step-by-step without Human Annotationscited by 928
  11. 2025Large Language Diffusion Modelscited by 912
  12. 2024Simple and Effective Masked Diffusion Language Modelscited by 848
  13. 2023Discrete Diffusion Modeling by Estimating the Ratios of the Data Distributioncited by 745
  14. 2023Muse: Text-To-Image Generation via Masked Generative Transformerscited by 648
  15. 2023Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Modelscited by 574
  16. 2022DiffuSeq: Sequence to Sequence Text Generation with Diffusion Modelscited by 450
  17. 2023Language Models can Solve Computer Taskscited by 372
  18. 2023How Language Model Hallucinations Can Snowballcited by 311
  19. 2023Do Large Language Models Know What They Don't Know?cited by 243
  20. 2024Scaling up Masked Diffusion Models on Textcited by 199
  21. 2024Taming Overconfidence in LLMs: Reward Calibration in RLHFcited by 112
  22. 2025A Survey on Diffusion Language Modelscited by 78
  23. 2024Rethinking the Spatial Inconsistency in Classifier-Free Diffusion Guidancecited by 59
  24. 2020Limitations of Autoregressive Models and Their Alternativescited by 55
  25. 2025ReVISE: Learning to Refine at Test-Time via Intrinsic Self-Verificationcited by 35
  26. 2024SELF-[IN]CORRECT: LLMs Struggle with Discriminating Self-Generated Responsescited by 26
  27. 2023How susceptible are LLMs to Logical Fallacies?cited by 23
  28. 2025Classifier-free Guidance with Adaptive Scalingcited by 22
  29. 2025RefineBench: Evaluating Refinement Capability of Language Models via Checklistscited by 15
  30. 2025In-situ Autoguidance: Eliciting Self-Correction in Diffusion Modelscited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025EDIT: Early Diffusion Inference Termination for dLLMs Based on Dynamics of Training Gradients
  2. 2026The Confidence Shortcut: A Reasoning Failure Mode of Masked Diffusion Models
  3. 2026Semantic Reasoning Denoising: Correcting Language Model Reasoning with Semantic Operators
  4. 2026ReTreVal: Reasoning Tree with Validation and Cross-Problem Memory for Large Language Models
  5. 2026Divergent-Convergent Reasoning: Scaling Test-Time Compute through Structured Solution Synthesis
  6. 2025Rethinking Reasoning with MDLMs: Early Exits, Post-hoc Reasoning, and Beyond
  7. 2026Test-Time Scaling with Diffusion Language Models via Reward-Guided Stitching
  8. 2025On the Reasoning Abilities of Masked Diffusion Language Models