Metraipolis › archive › dc_0198afbd9bea

The Ball and the Box: Two Geometries of Computation in Superposition — arXiv:2610.11744

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.11744
License (as archived)
CC0-1.0
Archived
2026-10-09T06:23:29.408Z
SHA-256 of the archived text
c6428b4c777d0817e616b8b4d6f2549fcd9c0134c592200bc8452d63940fe5c0
Stable URL of this page
https://metraipolis.com/documento/dc_0198afbd9bea

How to cite

The Ball and the Box: Two Geometries of Computation in Superposition — arXiv:2610.11744 — arXiv (en), https://arxiv.org/abs/2610.11744 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0198afbd9bea (Metraipolis archive dc_0198afbd9bea, sha256 c6428b4c777d0817e616b8b4d6f2549fcd9c0134c592200bc8452d63940fe5c0)
BibTeX
@misc{dc0198afbd9bea,
  title = {The Ball and the Box: Two Geometries of Computation in Superposition — arXiv:2610.11744},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.11744},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0198afbd9bea; sha256 c6428b4c777d0817e616b8b4d6f2549fcd9c0134c592200bc8452d63940fe5c0},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.11744. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Neural representations can encode more features than they have dimensions, a phenomenon known as superposition. We study the dimension needed to compute Boolean gates from such representations. For a single threshold layer with a Gaussian random dictionary and uniformly random sparse Boolean inputs, we derive sharp dimension thresholds under two error criteria. A vanishing expected error count can require more dimensions than correctness of every output with high probability. Shared reads explain the gap: rare realizations can produce many errors at once. The expected-count threshold has ball geometry, while joint reliability has box geometry when a gate is evaluated on every feature tuple. Optimizing shared readout weights and biases gives explicit thresholds for conjunction, disjunction, and majority. For pairwise conjunction, the analysis also describes the transition near the threshold, in agreement with exact simulations.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.11744 — 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 2026-10-09T05:50:31.603Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).