Temporal Self-Distillation: Faster Inference in Discrete Diffusion Language Models

• URL: http://arxiv.org/abs/2609.15177v1
• Date: Mon, 14 Sep 2026 07:59:45 GMT
• Title: Temporal Self-Distillation: Faster Inference in Discrete Diffusion Language Models
• Abstract summary: Diffusion language models (dLLMs) promise fast inference by generating multiple tokens in parallel, but suffer severe performance degradation when parallel decoding is pushed too aggressively.<n>We introduce Temporal Self-Distillation (TSD), a simple on-policy method that trains dLLMs for fast inference by distilling predictions across time.
• Score: 11.929379448468568
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Diffusion language models (dLLMs) promise fast inference by generating multiple tokens in parallel, but suffer severe performance degradation when parallel decoding is pushed too aggressively. We introduce Temporal Self-Distillation (TSD), a simple on-policy method that trains dLLMs for fast inference by distilling predictions across time. Specifically, TSD distills the model's denoising distribution at earlier timesteps toward its distribution at the final timestep at which a token is committed. This encourages earlier predictions to better anticipate the model's eventual output, enabling much more aggressive parallel decoding. Because its teacher signal comes from the model itself, TSD requires no offline teacher generation and applies seamlessly to both base and post-trained policies. Across seven benchmarks in mathematics, planning, and code, TSD substantially shifts the speed--quality frontier toward the low-compute regime. TSD thus provides a simple, single-stage approach to accelerating dLLMs, achieving speedups competitive with offline distillation while avoiding a complex two-stage pipeline.

Related papers

• DiffusionGemma Technical Report [72.10570774996219]
  We introduce DiffusionGemma, a language model that uses discrete diffusion to generate text at exceptionally high speed.<n>Rather than decoding one token at a time, DiffusionGemma iteratively refines blocks of 256 tokens in parallel.<n>Our compute-efficient two-stage training pipeline uses fewer than 10% of the starting AR model's total training token budget.
  arXiv  Detail & Related papers  (2026-07-31T16:11:46Z)
• Parallel Decoding Distillation for Fast Image and Video Generation [35.18845060233149]
  We introduce Parallel Decoding Distillation (PDD) for fast inference of diffusion and flow matching models.<n>PDD accelerates generation by predicting multiple denoising steps per network evaluation.<n>Our method achieves SOTA performance with 4-8 NFE on LTX-2.3 Text-to-Video/Audio, Wan 14B Text-to-Video, and Qwen-Image Text-to-Image.
  arXiv  Detail & Related papers  (2026-07-28T17:20:00Z)
• TAD: Temporal-Aware Trajectory Self-Distillation for Fast and Accurate Diffusion LLM [18.69783590125584]
  Diffusion large language models (dLLMs) offer a promising paradigm for parallel text generation.<n>They face an accuracy-parallelism trade-off, where increasing tokens per forward (TPF) often degrades generation quality.<n>Existing acceleration methods often gain speed at the cost of accuracy.<n>We propose TAD, a Temporal-Aware trajectory self-Distillation framework.
  arXiv  Detail & Related papers  (2026-05-10T13:38:53Z)
• d3LLM: Ultra-Fast Diffusion LLM using Pseudo-Trajectory Distillation [31.922313594074925]
  Diffusion large language models (dLLMs) offer capabilities beyond those of autoregressive (AR) LLMs.<n>Current methods typically focus on only one-side of the coin, targeting either efficiency or performance.<n>We propose d3LLM (Pseudo-Distilled Diffusion Large Language Model), striking a balance between accuracy and parallelism.
  arXiv  Detail & Related papers  (2026-01-12T14:25:36Z)
• CD4LM: Consistency Distillation and aDaptive Decoding for Diffusion Language Models [27.070045950001532]
  CD4LM is a framework that decouples training from inference.<n>On GSM8K, CD4LM matches the LLaDA baseline with a 5.18x wall-clock speedup.
  arXiv  Detail & Related papers  (2026-01-05T16:09:22Z)
• Fast and Accurate Causal Parallel Decoding using Jacobi Forcing [41.89066334075016]
  Jacobi Forcing is a progressive distillation paradigm where models are trained on their own generated parallel decoding trajectories.<n>We introduce multi-block decoding with rejection recycling, which enables up to 4.5x higher token acceptance count per iteration and nearly 4.0x wall-clock speedup.
  arXiv  Detail & Related papers  (2025-12-16T18:45:18Z)
• dParallel: Learnable Parallel Decoding for dLLMs [77.24184219948337]
  Diffusion large language models (dLLMs) offer parallel token prediction and lower inference latency.<n>Existing open-source models still require nearly token-length decoding steps to ensure performance.<n>We introduce dParallel, a simple and effective method that unlocks the inherent parallelism of dLLMs for fast sampling.
  arXiv  Detail & Related papers  (2025-09-30T16:32:52Z)
• Diffusion Language Models Know the Answer Before Decoding [56.96815863705218]
  Diffusion language models (DLMs) have emerged as an alternative to autoregressive approaches.<n>Our work highlights and leverage an overlooked property of DLMs early answer convergence.<n>We introduce Prophet, a training-free fast decoding paradigm that enables early commit decoding.
  arXiv  Detail & Related papers  (2025-08-27T15:40:25Z)
• Accelerating Diffusion LLMs via Adaptive Parallel Decoding [60.407727995313074]
  We introduce adaptive parallel decoding (APD), a novel method that dynamically adjusts the number of tokens sampled in parallel.<n>APD provides markedly higher throughput with minimal quality degradations on downstream benchmarks.
  arXiv  Detail & Related papers  (2025-05-31T06:10:10Z)
• Self-Refining Diffusion Samplers: Enabling Parallelization via Parareal Iterations [53.180374639531145]
  Self-Refining Diffusion Samplers (SRDS) retain sample quality and can improve latency at the cost of additional parallel compute.<n>We take inspiration from the Parareal algorithm, a popular numerical method for parallel-in-time integration of differential equations.
  arXiv  Detail & Related papers  (2024-12-11T11:08:09Z)
• Non-autoregressive Sequence-to-Sequence Vision-Language Models [59.445765313094434]
  We propose a parallel decoding sequence-to-sequence vision-language model that marginalizes over multiple inference paths in the decoder.<n>The model achieves performance on-par with its state-of-the-art autoregressive counterpart, but is faster at inference time.
  arXiv  Detail & Related papers  (2024-03-04T17:34:59Z)
• SpeedUpNet: A Plug-and-Play Adapter Network for Accelerating Text-to-Image Diffusion Models [4.232258232565518]
  Text-to-image diffusion models (SD) exhibit significant advancements while requiring extensive computational resources. LCM-LoRA, trainable once for diverse models, offers universality but rarely considers ensuring the consistency of generated content before and after acceleration. This paper proposes SpeedUpNet (SUN), an innovative acceleration module, to address the challenges of universality and consistency.
  arXiv  Detail & Related papers  (2023-12-13T09:42:04Z)

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.