Metraipolis › archive › dc_26935447b53e
EMAN: Optimization-Driven Capacity Growth through Path Emergence in Multi-Task Learning — arXiv:2608.16930
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.16930
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-18T14:29:05.062Z
- SHA-256 of the archived text
d9afc63fe5bf6b472347a35c482ad229c61775336a7b3626d0b38fb536c924d4- Stable URL of this page
https://metraipolis.com/documento/dc_26935447b53e
How to cite
EMAN: Optimization-Driven Capacity Growth through Path Emergence in Multi-Task Learning — arXiv:2608.16930 — arXiv (en), https://arxiv.org/abs/2608.16930 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_26935447b53e (Metraipolis archive dc_26935447b53e, sha256 d9afc63fe5bf6b472347a35c482ad229c61775336a7b3626d0b38fb536c924d4)
BibTeX
@misc{dc26935447b53e,
title = {EMAN: Optimization-Driven Capacity Growth through Path Emergence in Multi-Task Learning — arXiv:2608.16930},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.16930},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_26935447b53e; sha256 d9afc63fe5bf6b472347a35c482ad229c61775336a7b3626d0b38fb536c924d4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.16930. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Existing multi-task learning methods rely on hard sharing, multiple paths or experts, adaptive sharing, and dynamic expansion. However, their capacity changes are usually constrained by predefined structures or triggered by task boundaries and conflict signals. This raises a fundamental question: can a network start from exact single-path computation and grow a new independent path only when persistent optimization evidence appears? We propose the Emergent Modular Atomic Network (EMAN), an optimization-driven framework for exposing an antisymmetric growth direction through latent relative phases without instantiating a second path, and for monitoring multiple decision signals during training to transform local optimization evidence into a structural decision. EMAN materializes two equal-capacity independent paths only after certification. EMAN adaptively allocates shared and task-specific representation capacity to accommodate varying task requirements. Extensive experiments on controlled rank settings, PASCAL-Context, and NYUv2 validate its effectiveness, achieving improved performance at a competitive computational cost. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-19 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.16930 — 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 2026-10-18T14:27:54.355Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-19?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).