Metraipolis › archive › dc_a2708d5977cd
A Chosen Future Can Still Be Rewritten: Causal Writability in Video Models — arXiv:2609.15980
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.15980
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-16T00:12:13.289Z
- SHA-256 of the archived text
82897583fb73ad759daa7d820531218889fbc326b2e3cea65db52ea08d0a9459- Stable URL of this page
https://metraipolis.com/documento/dc_a2708d5977cd
How to cite
A Chosen Future Can Still Be Rewritten: Causal Writability in Video Models — arXiv:2609.15980 — arXiv (en), https://arxiv.org/abs/2609.15980 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_a2708d5977cd (Metraipolis archive dc_a2708d5977cd, sha256 82897583fb73ad759daa7d820531218889fbc326b2e3cea65db52ea08d0a9459)
BibTeX
@misc{dca2708d5977cd,
title = {A Chosen Future Can Still Be Rewritten: Causal Writability in Video Models — arXiv:2609.15980},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.15980},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_a2708d5977cd; sha256 82897583fb73ad759daa7d820531218889fbc326b2e3cea65db52ea08d0a9459},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.15980. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
When a video model generates physically incorrect motion, did it fail to learn the correct motion, or did it learn it but fail to use it? We show the latter: the correct motion remains available inside the model and can still be made to control the generated video. We train on videos where red masses oscillate slowly and blue masses oscillate quickly, then test a red mass with fast observed motion. Even when the model generates slow motion in this conflicting case, a low-dimensional edit predicted from simple physical variables restores the correct fast motion. We call this ability causal writability. At fixed strength, we find a sharp depth boundary: the same edit changes the video before the boundary but not after it. This closure marks commitment for that write. The motion signal nevertheless remains, and a stronger downstream write can restore physical motion, while excessive gain overshoots. Early causal writability predicts which errors training later corrects: those errors are writable at more network depths than errors that persist. We reproduce both causal writability and its sharp closure in a pretrained 1.3B video model, supporting generality across model scale and training regime. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-15 in cs.LG). Categorie arXiv: Machine Learning (cs.LG) ; Computer Vision and Pattern Recognition (cs.CV). Pagina del lavoro: https://arxiv.org/abs/2609.15980 — 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 2026-10-15T23:54:14.804Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).