CORTEXA
← Browse
arxivcs.LG2026-06-28

Adaptive Block Diffusion: Resolving Training-Inference Mismatch in Diffusion Language Models

Gagan Jain

Diffusion Language Models (DLMs) are typically trained under fixed context structures, restricting denoising to predetermined token subsets. This creates a mismatch between training and inference, where models must operate over arbitrary configurations, leading to degradation off the training grid. We propose Adaptive Block Diffusion (ABD), which resolves this mismatch by optimizing denoising risk over a distribution of prefix-window configurations. By treating the configuration as a stochastic variable, ABD trains a single model over the full configuration space without architectural changes. We show that generalization across decoding strategies is governed by the support of the training distribution, and that ABD guarantees denoising optimality for any inference policy whose configurations are covered during training. Empirically, ABD exhibits structural invariance across decoding scales, avoiding off-grid collapse and recovering a monotonic relationship between block size and perplexity, while matching or outperforming fixed-block specialists at their target scales.

View free PDFSource page

Related papers

arxivcs.LGcs.AIcs.CL2026-07-14

Accelerating Masked Diffusion Large Language Models: A Survey of Efficient Inference Techniques

Daehoon Gwak, Minhyung Lee, Junwoo Park, Jaegul Choo

Diffusion large language models (dLLMs) offer a theoretical advantage in parallel generation over standard autoregressive models. However, parallel generation alone does not guarantee practical speedups. Realizing this efficiency requires specialized inference mechanisms, such as…

View free PDFSource page
arxivcs.CLcs.AIcs.LG2026-07-18

Trace-Based On-Policy Distillation for Masked Diffusion Language Models

Haolin Ren, Ziyang Huang, Chenhao Yuan, Jun Zhao, Kang Liu

Diffusion large language models (dLLMs) are a promising alternative to autoregressive generation. However, reasoning-oriented post-training for dLLMs remains challenging. Supervised fine-tuning (SFT) for dLLMs requires dense but often off-policy masked states, while reinforcement…

View free PDFSource page
arxivcs.CLcs.LG2026-07-17

Adaptive Multi-Step Lookahead Decoding for Diffusion Language Models

Yingqian Cui, Wei Deng, Lantao Mei, Hang Li, Charu C. Aggarwal, Hui Liu, et al.

Masked diffusion language models (DLMs) enable parallel text generation by iteratively refining masked tokens, offering a promising alternative to autoregressive decoding. Recent lookahead-based decoding methods improve the accuracy--efficiency trade-off by exploring future decod…

View free PDFSource page
arxivcs.CLcs.LG2026-07-15

Graded Entity-Familiarity Readouts in Language Models: Polish Adaptation, Cross-Language Robustness, and Refusal Steering

Grzegorz Brzezinka

Can a language model estimate its familiarity with an entity before generating an answer? We study activations at the final prompt token in twelve instruction-tuned models from the Bielik, PLLuM, Gemma-4, and Qwen3 families, using a new dataset of 1,440 Polish entities spanning f…

View free PDFSource page
arxivcs.LGcs.CV2026-07-07

FourTune: Towards Fully 4-Bit Efficient Post-Training for Diffusion Models

Bowen Xue, Zihan Min, Xingyang Li, Zhekai Zhang, Haocheng Xi, Lvmin Zhang, et al.

Diffusion models have become a dominant paradigm for high-quality generative modeling, while post-training is essential for adapting them to diverse downstream applications. However, post-training of large diffusion models is still challenging due to the prohibitive memory footpr…

View free PDFSource page