Metraipolis › archive › dc_7c5486aefde2

MultiEcho: An Experimental Science of Learned Worlds — arXiv:2609.33347

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.33347
License (as archived): CC0-1.0
Archived: 2026-10-07T22:37:23.418Z
SHA-256 of the archived text: bef6797628d7715ff55ee42365d5b2ebef0de89898111fe10e826c1268453932
Stable URL of this page: https://metraipolis.com/documento/dc_7c5486aefde2

How to cite

MultiEcho: An Experimental Science of Learned Worlds — arXiv:2609.33347 — arXiv (en), https://arxiv.org/abs/2609.33347 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_7c5486aefde2 (Metraipolis archive dc_7c5486aefde2, sha256 bef6797628d7715ff55ee42365d5b2ebef0de89898111fe10e826c1268453932)

BibTeX

@misc{dc7c5486aefde2,
  title = {MultiEcho: An Experimental Science of Learned Worlds — arXiv:2609.33347},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.33347},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_7c5486aefde2; sha256 bef6797628d7715ff55ee42365d5b2ebef0de89898111fe10e826c1268453932},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.33347. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

World models can be studied as experimental systems with response laws of their own. We introduce MultiEcho, a framework for estimating these laws through controlled counterfactual interventions, delimiting their applicability, and separately testing their physical correspondence. Across nine simulated physical systems and seven frozen model configurations, three-reference estimators predict complete intervention responses and recover intervention parameters. Estimator selection uses discovery data only; frozen fits are evaluated on validation and confirmation contexts. The experiments distinguish response predictability, intervention readability and physical accuracy. Responses can be locally describable yet poorly match physical effects in the same target coordinates. Event-window, visibility and camera interventions reveal conditional applicability, and paired generator configurations show reduced readability under a scene prompt with stronger guidance. Magnitude sweeps expose small image errors alongside large relative effect errors. An exact-reset material experiment separates registered visible-response success from fixed-readout failure on material-dependent futures at matched positions and velocities. Exact finite-scale identities resolve odd and even response errors; first-order remainder bounds specify when refined calibration converges. MultiEcho provides an experimental basis for studying learned-world laws independently of, and in relation to, physical laws.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.33347 — 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 2026-10-07T22:14:27.925Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).