ABC: A ny-Subset Autoregression via Non-Markovian Diffusion B ridges in C ontinuous Time and Space

Gabe Guo, Thanawat Sornwanee, Lutong Hao, Elon Litman, Stefano Ermon, Jose Blanchet · 2026

Read ABC: A ny-Subset Autoregression via Non-Markovian Diffusion B ridges in C ontinuous Time and Space by Gabe Guo, Thanawat Sornwanee, Lutong Hao, Elon Litman, Stefano Ermon, Jose Blanchet (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-27443.txt.

Introduction

Guo and five colleagues present ABC, a method for making continuous-time, continuous-space random processes from partial views, such as first and last video frames. Older diffusion methods fail to capture how near-in-time states match, inject noise blind to time elapsed, and ignore any subsets of states. ABC models the process with a single ongoing stochastic differential equation whose clock and states track real time and process states. Tests show it beats rival methods on video generation and weather forecasting.

Catalog record

CREATORGabe Guo, Thanawat Sornwanee, Lutong Hao, Elon Litman, Stefano Ermon, Jose Blanchet
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE140329
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.27443

Browse the full library · About the Tower of Babel