PreprintarXiv licencearXiv
On physical implementability of reinforcement-learning-driven reasoning
Abstract
In the machine learning community, the reinforcement-learning-driven reasoning process has been intensively studied since around 2025. In this article, we present a negative result on the hardware implementation of this reinforcement-learning-driven reasoning process in a physical neural network model in the presence of a reward external potential energy for a given reward. In this hardware implementation, the existence of a chain of thought in dynamics contradicts the Lyapunov stability of the system.
Related
- Resource optimization for RIS-UAV integrated air-ground cooperative semantic communication driven by deep reinforcement learning
- Discovering like a scientist: Literature-informed dimensionless learning for physical closure of data-driven equations
- Deep Scientific Reasoning under Physical Constraints: Structure-Aware Spectrum Prediction
- Unitary implementability of Darboux transformations
- Reinforcement learning for robust calibration of multiqudit quantum gates
BibTeX
@misc{konishi2026physical,
title = {{On physical implementability of reinforcement-learning-driven reasoning}},
author = {Eiji Konishi},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.14131},
eprint = {2610.14131},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.14131},
}
RIS
TY - UNPB
TI - On physical implementability of reinforcement-learning-driven reasoning
AU - Eiji Konishi
PY - 2026
DA - 2026/10/09
PB - arXiv
DO - 10.48550/arxiv.2610.14131
UR - https://arxiv.org/abs/2610.14131
AB - In the machine learning community, the reinforcement-learning-driven reasoning process has been intensively studied since around 2025. In this article, we present a negative result on the hardware implementation of this reinforcement-learning-driven reasoning process in a physical neural network model in the presence of a reward external potential energy for a given reward. In this hardware implementation, the existence of a chain of thought in dynamics contradicts the Lyapunov stability of the system.
ER -