dUltra: Ultra-Fast Diffusion Language Models via Reinforcement Learning

• URL: http://arxiv.org/abs/2512.21446v1
• Date: Wed, 24 Dec 2025 23:31:48 GMT
• Title: dUltra: Ultra-Fast Diffusion Language Models via Reinforcement Learning
• Authors: Shirui Chen, Jiantao Jiao, Lillian J. Ratliff, Banghua Zhu,
• Abstract summary: Masked diffusion language models offer potential for parallel token generation.<n>Open-source MDLMs decode fewer than 5 tokens per model forward pass.<n>dUltra learns unmasking strategies for efficient parallel decoding.
• Score: 36.12942468805232
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Masked diffusion language models (MDLMs) offer the potential for parallel token generation, but most open-source MDLMs decode fewer than 5 tokens per model forward pass even with sophisticated sampling strategies. As a result, their sampling speeds are often comparable to AR + speculative decoding schemes, limiting their advantage over mainstream autoregressive approaches. Existing distillation-based accelerators (dParallel, d3LLM) finetune MDLMs on trajectories generated by a base model, which can become off-policy during finetuning and restrict performance to the quality of the base model's samples. We propose \texttt{dUltra}, an on-policy reinforcement learning framework based on Group Relative Policy Optimization (GRPO) that learns unmasking strategies for efficient parallel decoding. dUltra introduces an unmasking planner head that predicts per-token unmasking likelihoods under independent Bernoulli distributions. We jointly optimize the base diffusion LLM and the unmasking order planner using reward signals combining verifiable reward, distillation reward, and the number of unmasking steps. Across mathematical reasoning and code generation tasks, dUltra improves the accuracy--efficiency trade-off over state-of-the-art heuristic and distillation baselines, moving towards achieving ``diffusion supremacy'' over autoregressive models.

Related papers

• Self-Rewarding Sequential Monte Carlo for Masked Diffusion Language Models [58.946955321428845]
  This work presents self-rewarding sequential Monte Carlo (SMC)<n>Our algorithm stems from the observation that most existing MDLMs rely on a confidence-based sampling strategy.<n>We introduce the trajectory-level confidence as a self-rewarding signal for assigning particle importance weights.
  arXiv  Detail & Related papers  (2026-02-02T09:21:45Z)
• Learning Unmasking Policies for Diffusion Language Models [33.44995119635116]
  Language Models (dLLMs) now match the downstream performance of their autoregressive counterparts on many tasks.<n>One particularly successful variant is masked discrete diffusion, in which a buffer filled with special mask tokens is progressively replaced with tokens sampled from the model's vocabulary.<n>In this work, we propose to train sampling procedures using reinforcement learning.
  arXiv  Detail & Related papers  (2025-12-09T20:44:33Z)
• Principled RL for Diffusion LLMs Emerges from a Sequence-Level Perspective [85.06838178922791]
  Reinforcement Learning (RL) has proven highly effective for autoregressive language models.<n>But adapting these methods to diffusion large language models (dLLMs) presents fundamental challenges.<n>We propose a principled RL framework that treats entire sequence generation as a single action and uses the ELBO as a tractable sequence-level likelihood proxy.
  arXiv  Detail & Related papers  (2025-12-03T13:05:32Z)
• Lookahead Unmasking Elicits Accurate Decoding in Diffusion Language Models [51.12873073612084]
  Masked Diffusion Models (MDMs) as language models generate by iteratively unmasking tokens, yet their performance depends on the inference time order of unmasking.<n>We propose Lookahead Unmasking (LookUM), which addresses these concerns by reformulating sampling as path selection over all possible unmasking orders.<n>LookUM requires only two to three paths to achieve peak performance, demonstrating remarkably efficient path selection.
  arXiv  Detail & Related papers  (2025-11-04T02:37:37Z)
• Taming Masked Diffusion Language Models via Consistency Trajectory Reinforcement Learning with Fewer Decoding Step [28.12392773921128]
  Masked diffusion language models offer properties such as parallel decoding, flexible generation orders, and the potential for fewer inference steps.<n>A naive approach is to directly transfer techniques well-established for autoregressive (AR) language models to MDLMs.<n>We propose EOS Early Rejection (EOSER) and Ascending Step-Size (ASS) decoding schedulers, which unlock the potential of MDLMs to perform full diffusion-style decoding.
  arXiv  Detail & Related papers  (2025-09-28T15:01:15Z)
• Inference-Time Scaling of Diffusion Language Models with Particle Gibbs Sampling [70.8832906871441]
  We study how to steer generation toward desired rewards without retraining the models.<n>Prior methods typically resample or filter within a single denoising trajectory, optimizing rewards step-by-step without trajectory-level refinement.<n>We introduce particle Gibbs sampling for diffusion language models (PG-DLM), a novel inference-time algorithm enabling trajectory-level refinement while preserving generation perplexity.
  arXiv  Detail & Related papers  (2025-07-11T08:00:47Z)
• Plan for Speed: Dilated Scheduling for Masked Diffusion Language Models [13.575063025878208]
  Masked diffusion language models promise fast, non-autoregressive text generation.<n>Existing samplers, which pick tokens to unmask based on model confidence, ignore interactions when unmasking multiple positions in parallel.<n>We propose the Dilated Unmasking Scheduler (DUS), an inference-only, planner-model-free method that partitions sequence positions into non-adjacent dilated groups and unmasked them in parallel.
  arXiv  Detail & Related papers  (2025-06-23T18:49:23Z)
• 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)
• Reviving Any-Subset Autoregressive Models with Principled Parallel Sampling and Speculative Decoding [55.2480439325792]
  In arbitrary-order language models, it is an open question how to sample tokens in parallel from the correct joint distribution.<n>We find that a different class of models, any-subset autoregressive models (AS-ARMs), holds the solution.<n>We show that AS-ARMs achieve state-of-the-art performance among sub-200M parameter models on infilling benchmark tasks, and nearly match the performance of models 50X larger on code generation.
  arXiv  Detail & Related papers  (2025-04-29T06:33:13Z)

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.