Attention-Like Hebbian Learning from Quantum Probability Flow and Quantum-Annealer Tests
Journal of the Physical Society of Japan · 15.09.2026 · 95 (9)
DOI: 10.7566/jpsj.95.093001 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of the Physical Society of Japan, erschienen am 15.09.2026. Verfasst von Masayuki Ohzeki. Die Arbeit nennt 19 Literaturangaben. Gefördert von: Cabinet Office, Government of Japan.
Abstract (Originalsprache)
We propose a quantum probability-flow principle for deriving local learning rules in associative memory. A transverse field defines leakage channels from data states, and minimizing the measured survival loss gives stability-driven updates. For imaginary-time, dephased dynamics, the local leakage-free energy is the log-sum-exp of energy gaps; its gradient is a softmax-weighted Hebbian rule. Real-time stability instead yields a power-law weighting. D-Wave standard- and fast-anneal tests of a one-hot attention forward map are better fitted by an effective softmax than by a Lorentzian power law.
Förderung
- Cabinet Office, Government of Japan
Zitieren
APA
Ohzeki, M. (2026). Attention-Like Hebbian Learning from Quantum Probability Flow and Quantum-Annealer Tests. Journal of the Physical Society of Japan, 95(9). https://doi.org/10.7566/jpsj.95.093001
Vancouver
Ohzeki M. Attention-Like Hebbian Learning from Quantum Probability Flow and Quantum-Annealer Tests. Journal of the Physical Society of Japan. 2026;95(9). doi:10.7566/jpsj.95.093001
BibTeX
@article{Ohzeki2026,
author = {Ohzeki, Masayuki},
title = {Attention-Like Hebbian Learning from Quantum Probability Flow and Quantum-Annealer Tests},
journal = {Journal of the Physical Society of Japan},
year = {2026},
volume = {95},
number = {9},
doi = {10.7566/jpsj.95.093001}
}
EUVIMED