Metraipolis › archive › dc_6ce899ac622d
The Capability Manifold and ML Scaling Laws — arXiv:2609.27588
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.27588
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-08T09:12:48.218Z
- SHA-256 of the archived text
9a85f8111fcc3b02d39d8acd4ed1038bead95ef67845eb12ea6e739298f85c64- Stable URL of this page
https://metraipolis.com/documento/dc_6ce899ac622d
How to cite
The Capability Manifold and ML Scaling Laws — arXiv:2609.27588 — arXiv (en), https://arxiv.org/abs/2609.27588 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6ce899ac622d (Metraipolis archive dc_6ce899ac622d, sha256 9a85f8111fcc3b02d39d8acd4ed1038bead95ef67845eb12ea6e739298f85c64)
BibTeX
@misc{dc6ce899ac622d,
title = {The Capability Manifold and ML Scaling Laws — arXiv:2609.27588},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.27588},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6ce899ac622d; sha256 9a85f8111fcc3b02d39d8acd4ed1038bead95ef67845eb12ea6e739298f85c64},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.27588. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Existing machine learning (ML) scaling laws relate predictive loss to compute, model parameters, and data. However, as models are increasingly deployed through agentic harnesses, loss alone is insufficient to characterize downstream performance: models with similar loss can exhibit different capabilities in reasoning, retrieval, planning, and adaptation. Yet, no unified framework connects such capabilities to the coupled resources available across the ML lifecycle. We bridge this gap by introducing a capability manifold, a multidimensional framework mapping downstream capabilities to pre-training, post-training, and test-time resources through bounded scaling functions. Analytical Jacobians quantify capability sensitivity to resource changes and interactions. As an initial application, we embed Kaplan- and Chinchilla-type scaling laws and test-time compute within the framework, demonstrating how existing scaling relationships can be unified as trajectories on a common capability manifold. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-24 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2609.27588 — 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 2026-10-08T08:13:26.693Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-24?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).