Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Preregistration: Artifact-Injection Simulation Parameters, Paired-Delta QRNG Architecture

Andrea Rester Campbell

Published Aug 19, 2026• 0 citations
Publisher

Abstract

This preregistration fixes injection mechanisms, dose grid, and pass/fail criteria for a planned re-analysis of previously collected Baseline data from the Paired-Delta Protocol pilot (103 sessions, 3,090 blocks), ahead of manuscript resubmission. The two mechanisms (literal-bit composition bias, persistence) are drawn from NIST SP 800-90B's two mandated continuous health tests, not chosen freehand. The equivalence bound (δ=0.002) is the measured resolution floor of this dataset, derived from a whole-session bootstrap on the real, unperturbed data. Each dose will be classified as passing, indeterminate, or showing a detectable residual, for hit rate and the HRS ordering score.

DOI
10.17605/osf.io/bg9ys

Related papers

A Paired-Delta QRNG Architecture for Controlling Shared Artifacts in Observer–Randomness Experiments

Andrea Rester Campbell

Adds a real-perturbation QRNG provider-identity validation study to the manuscript (Abstract, Results, Limitations); result reported as indeterminate, not resolved. No changes to prior data, results, or conclusions.

•Dec 1, 2025•0 citations

TEC‑89: Ultra‑High‑Rigor Protocol for Intention Influence on QRNG

Vadim Streltsov

Protocol TEC‑89 implements the unified metrological framework of Protocol 9.0 to test the claimed phenomenon of intention influence on quantum random number generators (QRNG) – the alleged ability of a person (the "intender") to bias the output of a true hardware‑based quantum random number generator (e.g., ID Quantique Quantis, ANU Quantum Numbers, or equivalent) through intention alone, without physical contact. The primary observable is the deviation of the QRNG bit distribution (mean z‑score) from the expected chance level (50/50) during active intention periods, compared to control periods (no intention, sham intention, or trap conditions). The experimental design includes a certified hardware QRNG (Quantis‑v10.10.08 or equivalent) with direct USB connection (no buffer), real‑time visual/audio feedback for the intender (to enhance engagement), continuous logging of all bits with time stamps, triple blinding (intender, experimenter, and analyst are blind to condition), trap conditions (the intender is told to "influence" but the QRNG is replaced by a deterministic pseudo‑RNG), sham conditions (the intender is told to "influence" but is actually resting), catch conditions (a known bias is introduced into the QRNG output to validate the measurement system's sensitivity), positive controls (known effects, e.g., varying the number of bits per trial, to calibrate the system), physiological monitoring (skin conductance, heart rate variability, EEG alpha/theta power) to correlate with intention states, comprehensive environmental monitoring (temperature, humidity, electromagnetic fields), and automated data logging. The protocol mandates a pilot phase to calibrate the QRNG, establish the baseline noise floor, and validate the blinding procedure. The primary analysis uses a Linear Mixed Model (LMM) or a Binomial GLM comparing the z‑score (or proportion of 1s) between active and control conditions, with α = 0.025 (two‑tailed) and power ≥ 0.90 for a small effect (Cohen's d ≥ 0.05). The sample size is N = 10,000 bits per block (baseline, with adaptive increase to N = 100,000 bits if pilot shows high variability). Statistical analyses are preregistered, include Bayesian inference (BF₁₀), permutation tests, and multiverse robustness checks (≥70% of 12 forks). The aim is to detect a reproducible intention‑induced QRNG bias or to establish an upper bound on the effect size with unprecedented rigor.

•Aug 15, 2026•0 citations

Prospective Validation of Marginal, Differential, and Relational Diagnostics in the Paired-Delta Protocol

Andrea Rester Campbell

This project prospectively validates the Paired-Delta Protocol (PDP) as a diagnostic architecture for paired quantum random number generator (QRNG) data. PDP divides a single QRNG call into two simultaneously generated streams, allowing analysis of each stream individually, their paired difference, and statistical relationships between them. This expanded measurement space may provide diagnostic information unavailable in a conventional single-stream design, but it also requires validation that the architecture and its relational measures behave as intended under known conditions. The validation will test whether prespecified PDP diagnostics can distinguish known classes of data-generating changes. These include marginal changes affecting an individual stream, common-mode and differential changes, changes confined to the relationship between streams while preserving their marginal distributions, lagged relationships at prespecified lags from −2 through +2, and changes in the stability of local coupling that may occur without a corresponding change in average coupling. The validation will also assess whether alternative methods of partitioning the same QRNG output create systematic paired or relational structure. Diagnostic procedures, disturbance classes, parameter ranges, classification rules, and success criteria will be specified before generation of the held-out validation datasets. The validation data will then be generated under known ground-truth conditions and analyzed using the frozen diagnostic procedure. Performance will be evaluated by whether the procedure detects and correctly localizes the intended changes, preserves expected null and marginal properties, and distinguishes among the prespecified generating classes. Misclassification patterns will also be examined to identify diagnostic boundaries and mechanisms that cannot be distinguished from the paired observations alone. The purpose of this study is methodological. Developmental analyses used to construct and refine the diagnostic procedures are separate from the prospectively evaluated validation set and will not be counted as validation evidence. The expected outcome is a quantitative characterization of which forms of marginal, differential, temporal, and relational structure PDP can reliably distinguish under controlled conditions, together with explicit identification of its limitations. Successful validation would support the use of PDP's additional observables for exploratory hypothesis generation and for diagnostic characterization of independently established effects, rather than as additional opportunities for unconstrained significance testing.

•Aug 24, 2026•0 citations

A Preregistered Hardware Pilot of the NS1 Rate-Independence Protocol on a Superconducting Processor

Alicia Nastuk

SSAF's central falsifiable prediction, NS1, requires that under a witness-confirmed fixed classical context, the reduced dynamics of a probe carry no dependence on the quantum state: all fitted rate parameters are functions of the context alone. This record reports a preregistered pilot execution of the NS1 protocol on the superconducting processor ibm_fez, using target qubit 6 and independent witness qubit 10. The independent context witness passed its two-part gate. Raw stability gave p = 0.3617, and witness-rate homogeneity gave p = 0.9998. The primary endpoint found that one shared affine reduced-evolution law accounted for an informationally complete set of eight target preparations, with a leave-one-out statistic of 176.21 against a parametric-bootstrap null 95% quantile of 253.40 and a cross-validation p-value of 0.6167. The secondary endpoint found no preparation-dependent heterogeneity among six fitted transverse decay rates, with a homogeneity p-value of 0.4003 and a maximum fractional deviation of 2.5%. The result is evidence of consistency with the preregistered NS1 prediction at the stated experimental sensitivity. It is not unique confirmation of SSAF, and it does not exclude state dependence below the achieved resolution. Standard fixed-context open-system quantum mechanics also predicts consistency with this result. The deposited record includes the manuscript, TeX source, submitted Python and notebook protocol, preregistration and preflight records, raw and processed data, machine-readable results, figures, run provenance, licensing information, and checksums. The hardware run was executed on July 17, 2026, with 7,424 shots per effective circuit and 435.0 seconds of QPU usage.

•Jul 19, 2026•0 citations

Measurement as Record-Registration: A Locked-Null Protocol for Testing Record-Level Structure in Quantum Computing Readout

William T Partin

This manuscript presents a methods-forward, operational protocol for evaluating whether any structured dependence remains in quantum-computing measurement records after explicit modeling of standard quantum and device-level mechanisms. Classical definiteness is defined at the record layer, and pre-record states are treated as predictive descriptions rather than as classical outcome facts. Within this operational stance, the protocol converts questions about outcome status and record formation into a falsifiable test program grounded in preregistered analysis rules and standard measurement telemetry. A locked null-model family is specified in advance, incorporating standard quantum dynamics together with documented device-level mechanisms including drift, correlated and non-Markovian noise, crosstalk proxies, and leakage indicators. The framework provides operational definitions, a conservative postulate set, preregistered endpoints, residual-adjudication criteria, negative controls, a fixed windowing discipline, and a bound-first reporting standard designed to yield informative constraints under null results. Any candidate contribution beyond the declared null family must appear as structured, reproducible residual dependence that survives the locked null family and the full residual-filter sequence. Version 4.4 refines the protocol specification, clarifies evidentiary boundaries, and strengthens the integration of the synthetic demonstration. The manuscript includes a fully specified classical simulation that executes the decision pipeline under controlled conditions. The synthetic study illustrates how the protocol classifies a bounded-null case, a measured-artifact case, a partial-artifact/unresolved-covariate case, an injected structured-residual case, and a failed-topology case. This demonstration establishes computational executability of the protocol and clarifies the behavior of the decision tree, but it does not constitute empirical evidence for any physical mechanism. Because the protocol relies only on standard measurement telemetry and preregistered analysis rules, it is directly implementable across contemporary quantum-computing platforms and does not depend on any single dataset, backend, or hardware realization. The manuscript is a specification paper: it defines a reproducible evidentiary framework for record-level claims, while large-scale empirical implementation, hardware benchmarking, and cross-platform replication remain subsequent studies.

•Mar 17, 2026•0 citations

Preregistered Confirmatory Evaluation of State-Dependent Pre-Decoder Conditioning in Quantum Error Correction

Cody Robert Gleason

This project presents a preregistered, confirmatory evaluation of a state-dependent pre-decoder conditioning approach in quantum error correction (QEC). The study investigates whether an upstream intervention applied to syndrome-derived information—prior to decoding—produces a consistent and reproducible effect on downstream decoding behavior under controlled conditions. The central objective is to evaluate the existence, robustness, and generality of an upstream conditioning effect at the architectural level, without assuming or asserting a specific internal mechanism. The intervention is treated as a black-box transformation that modifies the representation of syndrome information before it is provided to a decoder, enabling assessment of its observable footprint in decoder-relevant outcomes. The experimental design follows a fully specified, preregistered protocol with all primary elements defined in advance. A four-arm structure is employed: (i) an untreated control condition, (ii) a uniformly applied intervention, (iii) a state-dependent differential intervention, and (iv) a budget-matched comparison that isolates allocation effects under fixed total intervention magnitude. All arms are evaluated across fixed code distances and physical error-rate grids using shared seed schedules and identical decoding backends. Evaluation is conducted in a decoder-agnostic framework in which conditioned representations are generated once per instance and routed unchanged to multiple downstream decoders. This ensures that any observed differences arise from the upstream intervention rather than decoder-specific adaptation or tuning. Primary outcomes are defined in terms of logical failure behavior and exposed error-load observables derived from each instance. These observables are combined into a standardized composite metric using control-only normalization procedures, with all transformations, binning rules, and statistical tests specified in advance and frozen prior to confirmatory analysis. Conditional analyses are performed using bins defined exclusively from control distributions, with consistent application across all experimental arms. The preregistration includes explicit governance rules covering pilot-to-confirmatory transition, parameter freeze conditions, and the handling of any deviations. Pilot-authorized adjustments (such as parameter grid refinement or binning resolution) are finalized and documented prior to continuation of confirmatory execution. No post hoc modification of primary endpoints, evaluation criteria, or aggregation procedures is permitted. The expected outcome of this study is not a predetermined direction of effect, but a rigorous determination of whether an upstream conditioning intervention produces a measurable, reproducible, and structurally consistent footprint in decoder-relevant behavior under preregistered conditions. Results will be reported strictly according to the predefined analysis plan, with negative controls and reproducibility checks included to validate robustness. This work is intended to establish or refute the presence of an architectural-level effect in QEC pipelines while maintaining a strict separation between publicly disclosed evaluation structure and non-public implementation details.

•Jan 1, 2026•0 citations
This preregistration fixes injection mechanisms, dose grid, and pass/fail criteria for a planned re-analysis of previously collected Baseline data from the Paired-Delta Protocol pilot (103 sessions, 3,090 blocks), ahead of manuscript resubmission. The two mechanisms (literal-bit composition bias, persistence) are drawn from NIST SP 800-90B's two mandated continuous health tests, not chosen freehand. The equivalence bound (δ=0.002) is the measured resolution floor of this dataset, derived from a whole-session bootstrap on the real, unperturbed data. Each dose will be classified as passing, indeterminate, or showing a detectable residual, for hit rate and the HRS ordering score.
Protocol TEC‑89 implements the unified metrological framework of Protocol 9.0 to test the claimed phenomenon of intention influence on quantum random number generators (QRNG) – the alleged ability of a person (the "intender") to bias the output of a true hardware‑based quantum random number generator (e.g., ID Quantique Quantis, ANU Quantum Numbers, or equivalent) through intention alone, without physical contact. The primary observable is the deviation of the QRNG bit distribution (mean z‑score) from the expected chance level (50/50) during active intention periods, compared to control periods (no intention, sham intention, or trap conditions). The experimental design includes a certified hardware QRNG (Quantis‑v10.10.08 or equivalent) with direct USB connection (no buffer), real‑time visual/audio feedback for the intender (to enhance engagement), continuous logging of all bits with time stamps, triple blinding (intender, experimenter, and analyst are blind to condition), trap conditions (the intender is told to "influence" but the QRNG is replaced by a deterministic pseudo‑RNG), sham conditions (the intender is told to "influence" but is actually resting), catch conditions (a known bias is introduced into the QRNG output to validate the measurement system's sensitivity), positive controls (known effects, e.g., varying the number of bits per trial, to calibrate the system), physiological monitoring (skin conductance, heart rate variability, EEG alpha/theta power) to correlate with intention states, comprehensive environmental monitoring (temperature, humidity, electromagnetic fields), and automated data logging. The protocol mandates a pilot phase to calibrate the QRNG, establish the baseline noise floor, and validate the blinding procedure. The primary analysis uses a Linear Mixed Model (LMM) or a Binomial GLM comparing the z‑score (or proportion of 1s) between active and control conditions, with α = 0.025 (two‑tailed) and power ≥ 0.90 for a small effect (Cohen's d ≥ 0.05). The sample size is N = 10,000 bits per block (baseline, with adaptive increase to N = 100,000 bits if pilot shows high variability). Statistical analyses are preregistered, include Bayesian inference (BF₁₀), permutation tests, and multiverse robustness checks (≥70% of 12 forks). The aim is to detect a reproducible intention‑induced QRNG bias or to establish an upper bound on the effect size with unprecedented rigor.
This project prospectively validates the Paired-Delta Protocol (PDP) as a diagnostic architecture for paired quantum random number generator (QRNG) data. PDP divides a single QRNG call into two simultaneously generated streams, allowing analysis of each stream individually, their paired difference, and statistical relationships between them. This expanded measurement space may provide diagnostic information unavailable in a conventional single-stream design, but it also requires validation that the architecture and its relational measures behave as intended under known conditions. The validation will test whether prespecified PDP diagnostics can distinguish known classes of data-generating changes. These include marginal changes affecting an individual stream, common-mode and differential changes, changes confined to the relationship between streams while preserving their marginal distributions, lagged relationships at prespecified lags from −2 through +2, and changes in the stability of local coupling that may occur without a corresponding change in average coupling. The validation will also assess whether alternative methods of partitioning the same QRNG output create systematic paired or relational structure. Diagnostic procedures, disturbance classes, parameter ranges, classification rules, and success criteria will be specified before generation of the held-out validation datasets. The validation data will then be generated under known ground-truth conditions and analyzed using the frozen diagnostic procedure. Performance will be evaluated by whether the procedure detects and correctly localizes the intended changes, preserves expected null and marginal properties, and distinguishes among the prespecified generating classes. Misclassification patterns will also be examined to identify diagnostic boundaries and mechanisms that cannot be distinguished from the paired observations alone. The purpose of this study is methodological. Developmental analyses used to construct and refine the diagnostic procedures are separate from the prospectively evaluated validation set and will not be counted as validation evidence. The expected outcome is a quantitative characterization of which forms of marginal, differential, temporal, and relational structure PDP can reliably distinguish under controlled conditions, together with explicit identification of its limitations. Successful validation would support the use of PDP's additional observables for exploratory hypothesis generation and for diagnostic characterization of independently established effects, rather than as additional opportunities for unconstrained significance testing.
SSAF's central falsifiable prediction, NS1, requires that under a witness-confirmed fixed classical context, the reduced dynamics of a probe carry no dependence on the quantum state: all fitted rate parameters are functions of the context alone. This record reports a preregistered pilot execution of the NS1 protocol on the superconducting processor ibm_fez, using target qubit 6 and independent witness qubit 10. The independent context witness passed its two-part gate. Raw stability gave p = 0.3617, and witness-rate homogeneity gave p = 0.9998. The primary endpoint found that one shared affine reduced-evolution law accounted for an informationally complete set of eight target preparations, with a leave-one-out statistic of 176.21 against a parametric-bootstrap null 95% quantile of 253.40 and a cross-validation p-value of 0.6167. The secondary endpoint found no preparation-dependent heterogeneity among six fitted transverse decay rates, with a homogeneity p-value of 0.4003 and a maximum fractional deviation of 2.5%. The result is evidence of consistency with the preregistered NS1 prediction at the stated experimental sensitivity. It is not unique confirmation of SSAF, and it does not exclude state dependence below the achieved resolution. Standard fixed-context open-system quantum mechanics also predicts consistency with this result. The deposited record includes the manuscript, TeX source, submitted Python and notebook protocol, preregistration and preflight records, raw and processed data, machine-readable results, figures, run provenance, licensing information, and checksums. The hardware run was executed on July 17, 2026, with 7,424 shots per effective circuit and 435.0 seconds of QPU usage.
This manuscript presents a methods-forward, operational protocol for evaluating whether any structured dependence remains in quantum-computing measurement records after explicit modeling of standard quantum and device-level mechanisms. Classical definiteness is defined at the record layer, and pre-record states are treated as predictive descriptions rather than as classical outcome facts. Within this operational stance, the protocol converts questions about outcome status and record formation into a falsifiable test program grounded in preregistered analysis rules and standard measurement telemetry. A locked null-model family is specified in advance, incorporating standard quantum dynamics together with documented device-level mechanisms including drift, correlated and non-Markovian noise, crosstalk proxies, and leakage indicators. The framework provides operational definitions, a conservative postulate set, preregistered endpoints, residual-adjudication criteria, negative controls, a fixed windowing discipline, and a bound-first reporting standard designed to yield informative constraints under null results. Any candidate contribution beyond the declared null family must appear as structured, reproducible residual dependence that survives the locked null family and the full residual-filter sequence. Version 4.4 refines the protocol specification, clarifies evidentiary boundaries, and strengthens the integration of the synthetic demonstration. The manuscript includes a fully specified classical simulation that executes the decision pipeline under controlled conditions. The synthetic study illustrates how the protocol classifies a bounded-null case, a measured-artifact case, a partial-artifact/unresolved-covariate case, an injected structured-residual case, and a failed-topology case. This demonstration establishes computational executability of the protocol and clarifies the behavior of the decision tree, but it does not constitute empirical evidence for any physical mechanism. Because the protocol relies only on standard measurement telemetry and preregistered analysis rules, it is directly implementable across contemporary quantum-computing platforms and does not depend on any single dataset, backend, or hardware realization. The manuscript is a specification paper: it defines a reproducible evidentiary framework for record-level claims, while large-scale empirical implementation, hardware benchmarking, and cross-platform replication remain subsequent studies.
This project presents a preregistered, confirmatory evaluation of a state-dependent pre-decoder conditioning approach in quantum error correction (QEC). The study investigates whether an upstream intervention applied to syndrome-derived information—prior to decoding—produces a consistent and reproducible effect on downstream decoding behavior under controlled conditions. The central objective is to evaluate the existence, robustness, and generality of an upstream conditioning effect at the architectural level, without assuming or asserting a specific internal mechanism. The intervention is treated as a black-box transformation that modifies the representation of syndrome information before it is provided to a decoder, enabling assessment of its observable footprint in decoder-relevant outcomes. The experimental design follows a fully specified, preregistered protocol with all primary elements defined in advance. A four-arm structure is employed: (i) an untreated control condition, (ii) a uniformly applied intervention, (iii) a state-dependent differential intervention, and (iv) a budget-matched comparison that isolates allocation effects under fixed total intervention magnitude. All arms are evaluated across fixed code distances and physical error-rate grids using shared seed schedules and identical decoding backends. Evaluation is conducted in a decoder-agnostic framework in which conditioned representations are generated once per instance and routed unchanged to multiple downstream decoders. This ensures that any observed differences arise from the upstream intervention rather than decoder-specific adaptation or tuning. Primary outcomes are defined in terms of logical failure behavior and exposed error-load observables derived from each instance. These observables are combined into a standardized composite metric using control-only normalization procedures, with all transformations, binning rules, and statistical tests specified in advance and frozen prior to confirmatory analysis. Conditional analyses are performed using bins defined exclusively from control distributions, with consistent application across all experimental arms. The preregistration includes explicit governance rules covering pilot-to-confirmatory transition, parameter freeze conditions, and the handling of any deviations. Pilot-authorized adjustments (such as parameter grid refinement or binning resolution) are finalized and documented prior to continuation of confirmatory execution. No post hoc modification of primary endpoints, evaluation criteria, or aggregation procedures is permitted. The expected outcome of this study is not a predetermined direction of effect, but a rigorous determination of whether an upstream conditioning intervention produces a measurable, reproducible, and structurally consistent footprint in decoder-relevant behavior under preregistered conditions. Results will be reported strictly according to the predefined analysis plan, with negative controls and reproducibility checks included to validate robustness. This work is intended to establish or refute the presence of an architectural-level effect in QEC pipelines while maintaining a strict separation between publicly disclosed evaluation structure and non-public implementation details.