Auditing Representation-Induced Geometry in Acoustic-Emission Fracture Transients
Abstract
Fixed convolutional representations can induce structured geometry in time-frequency data even without task-specific or source-domain training. We audit a previously reported morphology-space analysis of two granite acoustic-emission experiments. The original cumulative angular-path endpoint is reproduced to within 2.1e-7 in the Archive-B/Archive-A ratio. However, the same directional contrast is obtained with ImageNet features and with three randomly initialized EfficientNet-B0 encoders: the preserved Aether representation gives a ratio of 1.411, ImageNet gives 1.361, and the random encoders give 1.480--1.503. We therefore withdraw the attribution of this contrast to interferometric pretraining. We also correct a reversal between archived OG1/OG3 file identifiers and the physical experiments. Phase randomization and temporal shuffling alter the preserved trajectory, but these responses are reinterpreted as pipeline-dependent diagnostics rather than evidence of a learned universal elastic morphology. The corrected result is narrower: multiple fixed convolutional maps expose a reproducible contrast between two acoustic-emission datasets. The signal statistics and architectural biases responsible for that contrast remain open questions.
Related
- Effects of waveguide tube geometry and surrounding medium conditions on acoustic emission signal attenuation
- Advances in UAV-Based Methane Emission Monitoring: Sensors, Sampling Geometry, and Quantification
- Global stability of the Atlantic overturning circulation: edge state, long transients and boundary crisis under CO2 forcing
- Multi-frequency seismo-acoustic measurements of synthetic ice with fluid inclusions: implications for seismic and through-ice acoustic estimation of sea-ice thickness
- DEMONet: Underwater acoustic target recognition based on multi-expert network and cross-temporal variational autoencoder
BibTeX
@misc{sanchezandreu2026auditing,
title = {{Auditing Representation-Induced Geometry in Acoustic-Emission Fracture Transients}},
author = {Jose Sanchez-Andreu},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2606.17061},
eprint = {2606.17061},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2606.17061},
}
RIS
TY - UNPB
TI - Auditing Representation-Induced Geometry in Acoustic-Emission Fracture Transients
AU - Jose Sanchez-Andreu
PY - 2026
DA - 2026/09/01
PB - arXiv
DO - 10.48550/arxiv.2606.17061
UR - https://arxiv.org/abs/2606.17061
AB - Fixed convolutional representations can induce structured geometry in time-frequency data even without task-specific or source-domain training. We audit a previously reported morphology-space analysis of two granite acoustic-emission experiments. The original cumulative angular-path endpoint is reproduced to within 2.1e-7 in the Archive-B/Archive-A ratio. However, the same directional contrast is obtained with ImageNet features and with three randomly initialized EfficientNet-B0 encoders: the preserved Aether representation gives a ratio of 1.411, ImageNet gives 1.361, and the random encoders give 1.480--1.503. We therefore withdraw the attribution of this contrast to interferometric pretraining. We also correct a reversal between archived OG1/OG3 file identifiers and the physical experiments. Phase randomization and temporal shuffling alter the preserved trajectory, but these responses are reinterpreted as pipeline-dependent diagnostics rather than evidence of a learned universal elastic morphology. The corrected result is narrower: multiple fixed convolutional maps expose a reproducible contrast between two acoustic-emission datasets. The signal statistics and architectural biases responsible for that contrast remain open questions.
ER -