arXiv Machine Learning
Certification cost of quantum models: measurement correlation, not parameter count
Read the original on arXiv Machine Learning →The Flow has not summarised this story yet — read it at arXiv Machine Learning.
Related stories
arXiv Machine Learning
Jun 3
Spectral Anatomy of Quantum Gaussian Process Kernels
arXiv:2605. 30952v2 Announce Type: replace Abstract: Two recent results have reshaped quantum Gaussian processes (QGPs).
arXiv Machine Learning
Sep 22
The Price of Self-Calibration: Exact Evidence Budgets and Manufactured Blind Sets in Adaptive Monitoring
The paper derives precise cost formulas for self‑calibrating monitors that adjust thresholds online to maintain a specified long‑run false‑alarm rate under arbitrary drift. It shows that the guarantee is an accounting identity, independent of the monitored signal, and provides exact evidence identities for both step and ramp drift scenarios, as well as an exact law for the fluctuation of the certificate’s own alarm rate. Additionally, it proves that any monitor designed to tolerate a drift class is blind to all faults in the difference of that class, identifying the blind set for speed‑bounded drift classes and quantifying power outside this set with a sharp Gaussian projection bound.
arXiv Machine Learning
Jun 19
Effective Dimension Governs Generalization in Quantum Kernel Vision Models
arXiv:2606. 20183v1 Announce Type: new Abstract: Recent quantum vision models-quantum vision transformers and quantum convolutional networks-report two striking but unexplained empirical phenomena: (i) ansatze with more, or more uniformly distributed, entanglement generalize better, and (ii) injecting quantum noise can improve test accuracy rather than degrade it.
arXiv Machine Learning
Sep 17
QEMScore: How Much Does the Measurement Add to Learned Quantum Error Mitigation?
The paper introduces QEMScore, a metric that compares learned quantum error mitigators to capacity‑matched controls that do not use measurement data. Using simulated circuits with exact ideal answers, the study finds that many mitigators gain little from measurement inputs, with a plain polynomial model often outperforming them. On real hardware data, however, measurement inputs can provide predictive benefits, highlighting that performance depends on representation and protocol specifics.
arXiv AI
Aug 18
Pricing the Risk of Runtime Compression: Anytime-Valid Admission and a Served-Output Law for Compressed Serving State
arXiv:2608. 15810v1 Announce Type: new Abstract: Runtime compression of serving state trades quality for capacity with no priced guarantee: systems adapt precision on load signals with no soundness statement, and certified approaches budget request-level risk by a union bound over a pre-declared event count.
arXiv AI
4d ago
Quantization Error Is Spectrally Flat: A Single Random Probe Is a Calibrated, Data-Free Sensitivity Estimator, with Application to Budget-Targeted Mixed-Precision Quantization
arXiv:2609.33923v2 Announce Type: replace-cross Abstract: A single random Gaussian probe gives an unbiased estimate of the squared Frobenius norm of a layer's quantization error. The estimator is wel...