mlsanityData to 14 Oct 2026

Scheduling Thoughts: Learning the Order of Thought in Diffusion Language Models

arXiv 2606.23567 ↗First posted 22 Jun 2026cs.LG cs.AIcited by 1

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

also in 1 more
Abstract (arXiv)

Masked diffusion language models decode by iteratively unmasking tokens, where the unmasking order defines an "order of thought" that strongly influences generation quality yet is typically chosen heuristically. We derive a tractable upper bound on the sequential decoding mismatch, measured by the Kullback-Leibler divergence and expressed in terms of the model's pathwise log-likelihood, with tightness under sufficient model expressivity. This bound induces a dense self-aware reward over ordered trajectories, casting order selection as a principled policy optimization problem with a frozen denoiser. We instantiate this idea as Self-Aware Scheduling (SAS), which learns a lightweight order policy using Group Relative Policy Optimization and applies seamlessly to both any-order and semi-autoregressive decoding. On Sudoku with 1B MDM, SAS improves puzzle accuracy from 82.0% (best heuristic schedule) to 91.8%, and reaches 97.5% with second-stage fine-tuning along learned trajectories. On mathematical reasoning with LLaDA-8B, SAS improves pass@1 on GSM8K from 64% to 76% and on MBPP from 39.5% to 41%, consistently matching or exceeding heuristic schedules across generation lengths and block sizes. Project page: https://jimmyxu123.github.io/SAS

Key facts. Cited by 1 paper in the map (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 51 references that are in the map. First posted on arXiv 22 Jun 2026.

Builds on

44 of 51 references.

Closest ancestors · 4

Same direction, newest first

  1. 2025Enhancing Reasoning for Diffusion LLMs via Distribution Matching Policy Optimization9 Oct
  2. 2025DiFFPO: Training Diffusion LLMs to Reason Fast and Furious via Reinforcement Learning2 Oct
  3. 2025wd1: Weighted Policy Optimization for Reasoning in Diffusion Language Models7 Jul
  4. 2025d1: Scaling Reasoning in Diffusion Large Language Models via Reinforcement Learning16 Apr

Other references · 40

  1. 2020Denoising Diffusion Probabilistic Modelscited by 21,076
  2. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  3. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  4. 2020Score-Based Generative Modeling through Stochastic Differential Equationscited by 9,122
  5. 2021Program Synthesis with Large Language Modelscited by 3,637
  6. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,228
  7. 2021Variational Diffusion Modelscited by 1,153
  8. 2025Large Language Diffusion Modelscited by 912
  9. 2023Discrete Diffusion Modeling by Estimating the Ratios of the Data Distributioncited by 745
  10. 2024Simplified and Generalized Masked Diffusion for Discrete Datacited by 626
  11. 2021Argmax Flows and Multinomial Diffusion: Learning Categorical Distributionscited by 574
  12. 2025Dream 7B: Diffusion Large Language Modelscited by 544
28 more
  1. 2024Your Absorbing Discrete Diffusion Secretly Models the Conditional Distributions of Clean Datacited by 381
  2. 2025Train for the Worst, Plan for the Best: Understanding Token Ordering in Masked Diffusionscited by 214
  3. 2021Autoregressive Diffusion Modelscited by 210
  4. 2024Masked Diffusion Models are Secretly Time-Agnostic Masked Models and Exploit Inaccurate Categorical Samplingcited by 208
  5. 2024Scaling up Masked Diffusion Models on Textcited by 199
  6. 2023A Reparameterized Discrete Diffusion Model for Text Generationcited by 152
  7. 2025Accelerated Sampling from Masked Diffusion Models via Entropy Bounded Unmaskingcited by 131
  8. 2019Insertion Transformer: Flexible Sequence Generation via Insertion Operationscited by 125
  9. 2023Parallel Sampling of Diffusion Modelscited by 112
  10. 2025Dimple: Discrete Diffusion Multimodal Large Language Model with Parallel Decodingcited by 101
  11. 2025Revolutionizing Reinforcement Learning Framework for Diffusion Large Language Modelscited by 88
  12. 2025Diffusion Language Models Know the Answer Before Decodingcited by 70
  13. 2019Non-Monotonic Sequential Text Generationcited by 55
  14. 2019Insertion-based Decoding with automatically Inferred Generation Ordercited by 43
  15. 2024Accelerating Diffusion Models with Parallel Sampling: Inference at Sub-Linear Time Complexitycited by 42
  16. 2025SPG: Sandwiched Policy Gradient for Masked Diffusion Language Modelscited by 41
  17. 2024textitJump Your Steps: Optimizing Sampling Schedule of Discrete Diffusion Modelscited by 40
  18. 2024Causal Language Modeling Can Elicit Search and Reasoning Capabilities on Logic Puzzlescited by 38
  19. 2025Time Is a Feature: Exploiting Temporal Dynamics in Diffusion Language Modelscited by 33
  20. 2019Text Infillingcited by 32
  21. 2025Learning Unmasking Policies for Diffusion Language Modelscited by 30
  22. 2025Learning-Order Autoregressive Models with Application to Molecular Graph Generationcited by 29
  23. 2025Improving Discrete Diffusion Unmasking Policies Beyond Explicit Reference Policiescited by 22
  24. 2025Empirical Analysis of Decoding Biases in Masked Diffusion Modelscited by 19
  25. 2018Beyond Error Propagation in Neural Machine Translation: Characteristics of Language Also Mattercited by 18
  26. 2025Lipschitz-Guided Design of Interpolation Schedules in Generative Modelscited by 17
  27. 2025Parallel Sampling from Masked Diffusion Models via Conditional Independence Testingcited by 13
  28. 2025Boundary-Guided Policy Optimization for Memory-efficient RL of Diffusion Large Language Modelscited by 9

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 1.

1 paper.

  1. 2026Entropy Can Flow, or It Can Guide. Be Entropy. LEDFlow: Introducing Entropy-guided Generation Order into Uniform Discrete Flowcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Parallelism and Generation Order in Masked Diffusion Language Models: Limits Today, Potential Tomorrow
  2. 2026Thinking Out of Order: When Output Order Stops Reflecting Reasoning Order in Diffusion Language Models
  3. 2026The Flexibility Trap: Rethinking the Value of Arbitrary Order in Diffusion Language Models
  4. 2026Locally Confident, Globally Stuck: The Quality-Exploration Dilemma in Diffusion Language Models
  5. 2026Mask-Aware Policy Gradients for Diffusion Language Models
  6. 2026Escaping Confidence Trap: Evolutionary Decoding for Mathematical Reasoning in Diffusion LLMs
  7. 2026Improving Diffusion Language Model Decoding through Joint Search in Generation Order and Token Space
  8. 2026Hierarchical Continuous Diffusion Language Models