Metraipolis › archive › dc_88d9eb4e1f4f

Complexity Induction: Compositional Generalization via Structured Training Distortion — arXiv:2608.21464

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.21464
License (as archived)
CC0-1.0
Archived
2026-10-18T07:42:15.817Z
SHA-256 of the archived text
928db2153b3f8b3980994e311ecca14159cbc1440487245a110beeebf48cd776
Stable URL of this page
https://metraipolis.com/documento/dc_88d9eb4e1f4f

How to cite

Complexity Induction: Compositional Generalization via Structured Training Distortion — arXiv:2608.21464 — arXiv (en), https://arxiv.org/abs/2608.21464 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_88d9eb4e1f4f (Metraipolis archive dc_88d9eb4e1f4f, sha256 928db2153b3f8b3980994e311ecca14159cbc1440487245a110beeebf48cd776)
BibTeX
@misc{dc88d9eb4e1f4f,
  title = {Complexity Induction: Compositional Generalization via Structured Training Distortion — arXiv:2608.21464},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.21464},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_88d9eb4e1f4f; sha256 928db2153b3f8b3980994e311ecca14159cbc1440487245a110beeebf48cd776},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.21464. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We demonstrate that structured distortion of training data - which we term complexity induction - can induce compositional generalization in a standard CNN classifier without architectural modification. Using synthetic images of colored geometric shapes, we encode classes as flat string labels (e.g., "red-circle") with no explicit attribute decomposition, and exclude certain color-shape combinations from training entirely. We apply two distortion methods derived from Jaccard string similarity between class names: mixed labels (soft target distributions encoding inter-class overlap) and expanded dataset (false training samples with structurally motivated incorrect labels). Both methods induce the ability to predict unseen class combinations, and act at different levels: mixed labels activate the classifier for unseen combinations by exploiting the CNN's natural embedding structure, while expanded training improves the embedding factorization itself. A control with random (unstructured) false labels confirms that the effect depends on the structure of the distortion, not on noise per se. These results suggest that structured complication of training signals can influence both the internal organization of learned representations and their compositional interpretation - a principle that may underlie the role of natural language in cognitive development.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-25 in cs.AI).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2608.21464 — 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 2026-10-18T07:37:52.814Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).