The paper introduces Clock Diffusion, a framework for semi‑autoregressive continuous diffusion language models that incorporates position‑dependent noise schedules, efficient training, and sampling algorithms. It presents two generation modes—block and sliding window—and defines ClockDLMs, a family of Gaussian models that achieve state‑of‑the‑art diffusion likelihoods on OpenWebText and outperform continuous baselines on GSM8K while matching or exceeding discrete diffusion models. The authors also propose Cache Grab, a set of efficient samplers that adapt accelerated inference techniques from discrete diffusion to further improve model quality and efficiency.
By Yair Schiff, Omer Belhasin, Roy Uziel, Matan Rusanovsky, Ran Zilberstein, Marianne Arriola, Gilad Turok, Guanghan Wang, Volodymyr Kuleshov, Michael Elad
arXiv:2606. 06007v1 Announce Type: new Abstract: Generating realistic synthetic sequential data is critical in real-world applications across operations research, finance, healthcare, energy systems, and scientific computing, where time-indexed observations are used for prediction, simulation, risk assessment, and data-driven decision-making.
By Haoyang Cao, Minshuo Chen, Yinbin Han, Renyuan Xu
arXiv:2605. 19805v2 Announce Type: replace-cross Abstract: Irregular multivariate time series impose a trade-off for long-horizon forecasting: discrete methods can distort temporal structure via re-gridding, while continuous-time models often require sequential solvers prone to drift.
By Zinuo You, Jin Zheng, John Cartlidge
arXiv:2602. 17706v2 Announce Type: replace Abstract: Diffusion models learn data distributions indirectly through denoising, making the difficulty of generative modeling closely tied to the dependency structure of data.
By Rongyao Cai, Yuxi Wan, Kexin Zhang, Ming Jin, Zhiqiang Ge, Qingsong Wen, Yong Liu
HALO introduces a hyperspherical VAE to constrain continuous latent representations to a fixed‑radius shell, stabilizing numerical fluctuations. It then employs a masked autoregressive model that balances parallel decoding with temporal correlation learning, reducing inference steps and improving stability. Experiments show HALO achieves state‑of‑the‑art generation performance with significantly better inference efficiency compared to existing baselines.
By Chunyi Hou, Xiangfei Qiu, Hanyin Cheng, Yutong Li, Bin Yang
arXiv:2604.27443v3 Announce Type: replace
Abstract: Generating continuous-time, continuous-space stochastic processes (e.g., videos, weather forecasts) conditioned on partial observations (e.g., firs...
By Gabe Guo, Thanawat Sornwanee, Lutong Hao, Elon Litman, Stefano Ermon, Jose Blanchet
arXiv:2608. 02799v1 Announce Type: cross Abstract: Score-based diffusion models are typically formulated using continuous-time stochastic differential equations and measure-theoretic stochastic calculus.
By Sunder Ram Krishnan
arXiv:2607. 01775v1 Announce Type: new Abstract: Discrete diffusion models have steadily improved in quality relative to autoregressive (AR) models.
By Marianne Arriola, Volodymyr Kuleshov
arXiv:2606. 18186v1 Announce Type: cross Abstract: Finite-dimensional (FD) diffusion policies exhibit temporal drift owing to discretization artifacts that degrade long-horizon performance (when deployed on physical systems).
By Lekan Molu
Generating realistic synthetic sequential data is critical in real-world applications across operations research, finance, healthcare, energy systems, and scientific computing, where time-indexed observations are used for prediction, simulation, risk assessment, and data-driven decision-making. While diffusion models have achieved remarkable success in generating static data, their direct extensions to sequential settings often fail to capture temporal dependence and information structure.
arXiv:2606. 02241v1 Announce Type: new Abstract: Is the uniform-state diffusion framework a more powerful paradigm for discrete diffusion?
By Justin Deschenaux, Caglar Gulcehre
arXiv:2609.37974v1 Announce Type: cross
Abstract: Masked diffusion models (MDMs) generate text by unmasking several tokens per step, but they are trained and sampled under different conditions. The m...
By Manuel Madeira, Amitis Shidani, Alice Bizeul, Victor Turrisi, Louis B\'ethune, Bhavika Devnani, Dan Busbridge, Pierre Ablin, Jo\~ao Monteiro