Metraipolis › archive › dc_ee5b7edd1afa
All You Need Is Non-Commutative Words — arXiv:2608.29314
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.29314
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T08:06:56.817Z
- SHA-256 of the archived text
71f7f1f5b3b026bae8ae17092ee9abb0b98371b198080491220794578f6033b0- Stable URL of this page
https://metraipolis.com/documento/dc_ee5b7edd1afa
How to cite
All You Need Is Non-Commutative Words — arXiv:2608.29314 — arXiv (en), https://arxiv.org/abs/2608.29314 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ee5b7edd1afa (Metraipolis archive dc_ee5b7edd1afa, sha256 71f7f1f5b3b026bae8ae17092ee9abb0b98371b198080491220794578f6033b0)
BibTeX
@misc{dcee5b7edd1afa,
title = {All You Need Is Non-Commutative Words — arXiv:2608.29314},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.29314},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ee5b7edd1afa; sha256 71f7f1f5b3b026bae8ae17092ee9abb0b98371b198080491220794578f6033b0},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.29314. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We represent lexical tokens as unitary matrices and encode each sentence as their ordered product. The noncommutativity of matrix product captures word order without positional encodings (PEs). The same algebra yields several capabilities, including antisymmetric self-attention with no query, key, or value projections, and parallel composition of variable-length text chunks at a reduced attention cost. Furthermore, it provides a canonical-coset readout layer that encodes all true unitary degrees of freedom compactly, while supporting continual learning through nested group extensions that enlarge the operator space with each new task preserving prior representations exactly. Across standard text-classification benchmarks, the method matches or exceeds bag-of-words baselines. Achieving higher accuracy on IMDB and comparable performance on AG News. Notably, this is accomplished by replacing the conventional $\sim$30,000-dimensional vocabulary space with a dense, 64-parameter real-valued encoding, highlighting the expressive efficiency of our parameterization. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-01 in cs.CL). Categorie arXiv: Computation and Language (cs.CL). Pagina del lavoro: https://arxiv.org/abs/2608.29314 — 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 2026-10-17T07:47:52.689Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-01?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).