mlsanityData to 14 Oct 2026

Thinking Inside the Mask: In-Place Prompting in Diffusion LLMs

arXiv 2508.10736 ↗First posted 14 Aug 2025cs.CLcited by 13

On the map: LLM reasoning → Prompting and chain-of-thought reasoning

also in 1 more
Abstract (arXiv)

Despite large language models (LLMs) have achieved remarkable success, their prefix-only prompting paradigm and sequential generation process offer limited flexibility for bidirectional information. Diffusion large language models (dLLMs) present new opportunities through their bidirectional attention mechanisms and iterative refinement processes, enabling more flexible in-place prompting strategies. We introduce ICE (In-Place Chain-of-Thought Prompting with Early Exit), a novel framework that transforms prefix-only prompting into in-place prompting specifically designed for dLLMs. ICE integrates in-place prompts directly within masked token positions during iterative refinement and employs a confidence-aware early exit mechanism to significantly reduce computational overhead. Extensive experiments demonstrate ICE's effectiveness, achieving up to 17.29% accuracy improvement with 4.12× speedup on GSM8K, and up to 276.67× acceleration on MMLU while maintaining competitive performance.

Key facts. Cited by 13 papers in the map (data through 14 Oct 2026). Placed in Prompting and chain-of-thought reasoning, one of 17 directions in LLM reasoning. Builds on 24 of its 33 references that are in the map. First posted on arXiv 14 Aug 2025.

Builds on

24 of 33 references.

References · 24

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,399
  2. 2023LLaMA: Open and Efficient Foundation Language Modelscited by 14,132
  3. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  4. 2020Measuring Massive Multitask Language Understandingcited by 7,963
  5. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  6. 2021Program Synthesis with Large Language Modelscited by 3,637
  7. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,879
  8. 2023A Survey of Large Language Modelscited by 2,576
  9. 2025Large Language Diffusion Modelscited by 912
  10. 2024Simple and Effective Masked Diffusion Language Modelscited by 848
  11. 2023Discrete Diffusion Modeling by Estimating the Ratios of the Data Distributioncited by 745
  12. 2024Simplified and Generalized Masked Diffusion for Discrete Datacited by 626
12 more
  1. 2025Fast-dLLM: Training-free Acceleration of Diffusion LLM by Enabling KV Cache and Parallel Decodingcited by 420
  2. 2023The Reversal Curse: LLMs trained on "A is B" fail to learn "B is A"cited by 407
  3. 2025MMaDA: Multimodal Large Diffusion Language Modelscited by 250
  4. 2025Mercury: Ultra-Fast Language Models Based on Diffusioncited by 235
  5. 2025dLLM-Cache: Accelerating Diffusion Large Language Models with Adaptive Cachingcited by 169
  6. 2025dKV-Cache: The Cache for Diffusion Language Modelscited by 143
  7. 2025FlashDLM: Accelerating Diffusion Language Model Inference via Efficient KV Caching and Guided Diffusioncited by 77
  8. 2025Accelerating Diffusion Large Language Models with SlowFast Sampling: The Three Golden Principlescited by 45
  9. 2024SepLLM: Accelerate Large Language Models by Compressing One Segment into One Separatorcited by 41
  10. 2025LongLLaDA: Unlocking Long Context Capabilities in Diffusion LLMscited by 40
  11. 2025Plan for Speed: Dilated Scheduling for Masked Diffusion Language Modelscited by 34
  12. 2025The Devil behind the mask: An emergent safety vulnerability of Diffusion LLMscited by 34

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 6; 2026: 7.

13 papers.

Sort:

By area:

  1. 2025From Next-Token to Next-Block: A Principled Adaptation Path for Diffusion LLMscited by 24
  2. 2025AudioMarathon: A Comprehensive Benchmark for Long-Context Audio Understanding and Efficiency in Audio LLMscited by 22
  3. 2025Soft-Masked Diffusion Language Modelscited by 16
  4. 2025DiffuGuard: How Intrinsic Safety is Lost and Found in Diffusion Large Language Modelscited by 12
  5. 2025Rainbow Padding: Mitigating Early Termination in Instruction-Tuned Diffusion LLMscited by 10
  6. 2026dTRPO: Trajectory Reduction in Policy Optimization of Diffusion Large Language Modelscited by 3
  7. 2026Plan, Verify and Fill: A Structured Parallel Decoding Approach for Diffusion Language Modelscited by 2
  8. 2026Self-Purification Mitigates Backdoors in Multimodal Diffusion Language Modelscited by 2
  9. 2026DIP: Dynamic In-Context Planner For Diffusion Language Modelscited by 1
  10. 2025STaRR: Spatial-Temporal Token-Dynamics-Aware Responsive Remasking for Diffusion Language Modelscited by 0
  11. 2026Progressive Refinement Regulation for Accelerating Diffusion Language Model Decodingcited by 0
  12. 2026When Confidence Misleads: Suffix Anchoring and Anchor-Proximity Confidence Modulation for Diffusion Language Modelscited by 0
1 more
  1. 2026In-Place Instruction Following in Diffusion Language Modelscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Beyond the Prompt in Large Language Models: Comprehension, In-Context Learning, and Chain-of-Thought
  2. 2024Efficient Prompting Methods for Large Language Models: A Survey
  3. 2026Prompt Optimization Via Diffusion Language Models
  4. 2025Unlocking the Potential of Diffusion Language Models through Template Infilling
  5. 2023Guiding Large Language Models via Directional Stimulus Prompting
  6. 2026The PICCO Framework for Large Language Model Prompting: A Taxonomy and Reference Architecture for Prompt Structure
  7. 2026Diffusion Language Models Are Natively Length-Aware
  8. 2025Cross-Lingual Prompt Steerability: Towards Accurate and Robust LLM Behavior across Languages