Skip to content
DraconiticResearch
Search Sources API
197,693 studies31 sourcesUpdated 24 min ago
PreprintLicence not statedarXiv (Cornell University)

The data geometry of masking diffusion: Certified-optimal schedules via unmasking growth complexity

  • Martin J. Wainwright1
  1. 1Lab for Information and Decision Systems Statistics and Data Science Center EECS and Mathematics , Massachusetts Institute of Technology
Open at sourcePDFDOICite

Related

  1. The information geometry of product-reference discrete diffusion: Interaction growth complexity and optimal scheduling
    Martin J. Wainwright
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  2. Serial Depth of Adaptive Parallel Sampling: Lower Bounds and Schedules for Masked Diffusion Models
    Chuling Wen, Weijie Liang, Jian Lu +1
    arXivPreprintarXiv licenceFull textStatisticsarXiv
  3. Certified Inference and Training for Deep Equilibrium Networks: A Continuation Framework with Polynomial Complexity Guarantees
    Alex Borisevich
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  4. Pay Only for Disagreement: Certified No-Regression Verdicts for Model Updates with Matching Label-Complexity Bounds
    Vishnu Bindu Balachandran
    arXivPreprintCC BY 4.0Full textStatisticsarXiv
  5. Algorithmic Information Dynamics of Learning: A Certified, Differentiable Complexity Controller for Grokking
    Luan Ozelim, Abicumaran Uthamacumaran, Hector Zenil
    arXivPreprintCC BY-NC-SA 4.0Full textGeneral mathematicsarXiv

Cite

BibTeXRIS
BibTeX
@misc{wainwright2026data,
  title = {{The data geometry of masking diffusion: Certified-optimal schedules via unmasking growth complexity}},
  author = {Martin J. Wainwright},
  howpublished = {arXiv (Cornell University)},
  year = {2026},
  month = aug,
  doi = {10.48550/arxiv.2608.13520},
  url = {https://arxiv.org/abs/2608.13520},
}
RIS
TY  - UNPB
TI  - The data geometry of masking diffusion: Certified-optimal schedules via unmasking growth complexity
AU  - Martin J. Wainwright
PY  - 2026
DA  - 2026/08/13
PB  - arXiv (Cornell University)
DO  - 10.48550/arxiv.2608.13520
UR  - https://arxiv.org/abs/2608.13520
SN  - 2331-8422
ER  -