Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Online Access: | https://doi.org/10.5281/zenodo.18897853 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <h2>Overview</h2> <p>This initial release provides the first comprehensive experimental and simulated validation of the <strong>Retrodiction Landauer Principle</strong> ($F \geq e^{-\Sigma/2}$) in the context of quantum error correction.</p> <h2>Key Contents</h2> <ul> <li><strong>Hardware Benchmarking:</strong> 100% satisfaction of the predicted bound across 1,020 data points from IBM (Brisbane, Sherbrooke, Torino) and Google (Sycamore, Willow) quantum processors.</li> <li><strong>Decoder Hierarchy Analysis:</strong> Verification of the 36/36 performance match across MWPM, Unweighted MWPM, and Greedy decoders using Google's <code>stim</code> simulator.</li> <li><strong>Thermodynamic Filtering:</strong> Confirmation of post-selection scaling ($\ln R = \alpha \cdot d + \beta$) across independent experimental datasets.</li> <li><strong>Optimal Recovery Proof:</strong> Theoretical and numerical evidence showing the Petz recovery map achieves the highest fidelity among all retrodiction-consistent maps.</li> </ul> <h2>Limitations & Scope</h2> <p>Note that $F$ and $\Sigma$ were computed from reported channel parameters rather than simultaneous direct measurements. This version focuses on the stationary-Petz-recovery regime for Markovian and non-Markovian noise.</p>