Metraipolis › archive › dc_020543728580
The information geometry of product-reference discrete diffusion: Interaction growth complexity and optimal scheduling — arXiv:2608.28949
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.28949
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T07:47:24.398Z
- SHA-256 of the archived text
df3a5b59635a16bc95b5f70965c3363f593a611af30e0e4077f76ee8c9fd1795- Stable URL of this page
https://metraipolis.com/documento/dc_020543728580
How to cite
The information geometry of product-reference discrete diffusion: Interaction growth complexity and optimal scheduling — arXiv:2608.28949 — arXiv (en), https://arxiv.org/abs/2608.28949 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_020543728580 (Metraipolis archive dc_020543728580, sha256 df3a5b59635a16bc95b5f70965c3363f593a611af30e0e4077f76ee8c9fd1795)
BibTeX
@misc{dc020543728580,
title = {The information geometry of product-reference discrete diffusion: Interaction growth complexity and optimal scheduling — arXiv:2608.28949},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.28949},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_020543728580; sha256 df3a5b59635a16bc95b5f70965c3363f593a611af30e0e4077f76ee8c9fd1795},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.28949. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We study a class of product-reference diffusion algorithms for sampling from a discrete distribution. We show that their sampling performance can be characterized using a path-based measure of data geometry that we call the interaction growth complexity (IGC). We show that a bivariate IGC kernel gives an exact representation of both the KL discretization error and a simple one-step upper bound. The simpler univariate IGC density can be used to study the effect of stepsize choices on the iteration complexity required to obtain $\epsilon$-accurate samples in KL divergence. Samplers that traverse the path with equi-spaced steps in log-squared-reliability-odds have performance that depends on the aggregate IGC mass, whereas refined choices of stepsizes have a lower complexity depending on a square-root functional. In the fine-grid limit, both of these characterizations become sharp. We also allow general product reference distributions and show that the reference law can substantially reshape the IGC profile and the resulting sampling complexity; in particular, references far from both the uniform and the data marginals can yield dimension-dependent improvements. Finally, the aggregate IGC mass admits bounds in terms of total correlation and dual total correlation, thereby connecting the pathwise geometry to classical measures of multivariate dependence. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.AI). Categorie arXiv: Machine Learning (stat.ML) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Statistics Theory (math.ST). Pagina del lavoro: https://arxiv.org/abs/2608.28949 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-17T07:28:26.705Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).