Learned quantum-battery controls link high stored energy with suppressed participation
Source check: Source checked. Not independent scientific confirmation.
Finding: Soft actor-critic learning selected time-dependent couplings in a simulated anisotropic Dicke quantum battery with Kerr nonlinearity and photon loss. Compared with fixed couplings, the learned controls sustained higher stored energy alongside a persistently lower participation ratio. This is a conditional computed dynamical relationship, not a measured battery or proof of a general charging advantage.
Technology: Soft actor-critic reinforcement learning
Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.
- AI contribution
- A learned policy generates the time-dependent controls whose simulated energy and participation-ratio trajectories support the scoped dynamical finding.
- Test and result
- The learned policy changes rotating and counter-rotating couplings in the specified dissipative model. Participation-ratio trajectories were compared with a fixed-coupling reference across three Kerr strengths.
- Limitation / still unknown
- The result depends on the Hamiltonian, initial state, dissipative model and numerical truncation. Optimized variable controls have a broader control space than the fixed-coupling baseline. The participation ratio is a dynamical diagnostic, not an independently established universal mechanism. Separate analytically computed phase-transition results are not credited to AI; no physical device or energy-efficiency benefit was measured.
Evidence sources and annotation context
Evidence context (descriptions follow the cited sources; test type is a catalog classification):
Evidence sources: https://arxiv.org/html/2610.28373v1 · https://arxiv.org/abs/2610.28373v1
Source checked · editorial details
The source page was checked against this entry. Independent confirmation is not assessed by this label.
Source type: paper. Paper: linked separately.
Review date: 21 Nov 2026. Review scope: Primary v1 front, physical model, RL ResultsIII.3, conclusions and computational convergence context.
No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.
Editorial assessment: Impact 1/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.