Qubit decoherence

Qubit decoherence

Superconducting-qubit decoherence provides a useful test case for the physics bridge because a single device exposes several experimentally distinct processes. Energy relaxation, transverse-coherence loss, phase-sensitive dephasing, readout error, and calibration drift are related, but they are not interchangeable observables.

In this analysis, I use those distinctions to examine a metric-first version of the Sense bridge. The objective is not to identify the Sense coordinates with qubit parameters or to propose a new model of decoherence. The objective is to determine whether source-linked measurements can constrain how the coordinates q_S, c_S, phi_S, mu_S, and tau_S are used in one local physical context.

Here, calibration has a specific meaning. No five-parameter fit is performed. Instead, each physical surface is allowed to narrow, split, or block a model-side interpretation. A useful bridge should preserve distinctions already required by the experiment and should abstain when the source does not support a stronger statement.

Why qubit decoherence is a discriminating test

The term decoherence is often used for the general loss of quantum-state quality. Experimentally, however, several processes must be separated.

A T1 experiment probes energy relaxation. The qubit is prepared in the excited state, allowed to evolve for a controlled delay, and measured. The resulting decay characterizes excited-state population retention under the declared protocol.

Ramsey and echo experiments probe transverse coherence. A Ramsey sequence is sensitive to detuning and slowly varying frequency noise. An echo sequence introduces a refocusing pulse and suppresses part of that low-frequency contribution. The two protocols therefore need not return the same coherence time.

Readout introduces another layer. The qubit state is coupled to a resonator, converted into an analog response, amplified, represented in an I/Q plane, and assigned a label by a calibrated decision rule. An assignment error is not automatically a state-decay event, and a well-separated readout cluster does not establish that phase coherence was preserved.

All of these quantities may also vary with operating point, acquisition time, pulse calibration, or device history. Drift can alter the interpretation of T1, Ramsey, echo, and readout data even when the experimental labels remain unchanged.

This combination makes a superconducting qubit an appropriate bridge object. The physical data already impose a typed structure: population retention, coherence, phase-sensitive loss, readout, and temporal context must remain distinguishable. If the Sense coordinates cannot respect that structure, the proposed correspondence has no physical content.

The relation to the Sense bridge

The local calibration map used in this route is

q_S → excited-state retention and T1-like surfaces

c_S → protocol-bound carrier coherence, including declared Ramsey or echo surfaces

phi_S → phase-sensitive structure, such as Ramsey–echo separation or a valid Tphi diagnostic

mu_S → readout, SPAM, I/Q calibration, and label assignment

tau_S → drift, run order, timing, and calibration history

These arrows denote calibration against physical metrics. They are not identities.

q_S is not T1.
c_S is not T2.
phi_S is not a qubit phase or a dephasing rate.
mu_S is not assignment fidelity.
tau_S is not laboratory time.

The expected contribution was therefore not a relabeling of familiar observables. I tested whether the mapping imposed non-mixing rules.

Readout error should not be absorbed into coherence loss. Drift should not be absorbed into q_S, c_S, or phi_S. A generic T2 label should not be promoted to Ramsey or echo without protocol evidence. A pure-dephasing quantity should not be constructed from incompatible T1 and T2 measurements. An ordered source axis should not be interpreted as physical time unless its unit is bound. If the readout shell or source identity is removed, the route should abstain from a population interpretation.

This is the operational connection to the bridge concept. The experiment supplies the metric surfaces; the Sense grammar proposes a local organization; the return audit determines which parts of that organization remain admissible.

Why these source corpora were selected

The source search was organized by physical role rather than by topic alone. A collection of qubit files is not sufficient if its T1, Ramsey, echo, readout, and drift surfaces refer to different devices or unrelated calibration windows.

The primary corpus was Zenodo record 11034817, the source-data package for “Mechanically Induced Correlated Errors on Superconducting Qubits with Relaxation Times Exceeding 0.4 Milliseconds.” The selected lane was the Fig. 1, qubit 3 family. It linked Rabi-amplitude calibration, T1, Ramsey, and echo notebooks to one source identity and one qubit lane, with additional I/Q and drift-related context.

This internal linkage was more important than the reported lifetime. The bridge required several physical surfaces from a common experimental context. Combining a T1 value from one qubit, an echo result from another, and a readout plot from a third would produce a numerically plausible but physically artificial object.

Three additional records were retained as independent support or stress cases.

Zenodo 8004359 provided T1/T2 material together with fidelity and QND-readout surfaces. It was useful for testing whether a generic T2 label would be overinterpreted as Ramsey, echo, or pure-dephasing evidence.

Zenodo 17312111 provided an independent T1/Ramsey/echo-type family and a separate temporal surface.

Zenodo 13961130 provided readout, confusion-style, and feedback-latency surfaces.

These records were not merged into the primary experiment. Their purpose was to test whether the same bridge rules remained meaningful under different source conventions without importing a missing quantity from another device or figure.

What the analysis was expected to recover

The expected result was a typed reinterpretation of the measurement chain.

For q_S, the relevant question was whether excited-state retention could be identified from a source-bound T1 protocol.

For c_S, the question was whether a declared coherence protocol could be associated with the carrier without replacing all T2-type measurements by one undifferentiated quantity.

For phi_S, the question was whether a phase-sensitive contribution could be separated from relaxation and from the choice of coherence protocol.

For mu_S, the question was whether the I/Q-to-label chain was sufficiently specified to support a population statement.

For tau_S, the question was whether dynamic delay, acquisition order, wall-clock chronology, and calibration drift were separately available.

The standard relation

Gamma_phi = 1/T2 - 1/(2T1)

was used only as a physics-side constraint under its usual assumptions. It is meaningful when T1 and the chosen T2 quantity have compatible units, device identity, operating point, protocol, fit model, and calibration context. It is not an equation for phi_S.

A positive result required more than a set of matching names. The route had to retain the physical separations and had to stop when one of the required source maps was absent.

What the source audit changed

The early audits identified two immediate limitations.

First, an ordered decay trace was available in a source slice whose horizontal coordinate was not authoritatively bound to a physical time unit. The trace supported ordering and decay shape, but not a unit-bearing lifetime. This distinction forced tau_S to separate at least four objects: a source-axis coordinate, a physical delay, an acquisition chronology, and a calibration history.

Second, a T2-like surface in another source slice was not fully bound to a specific Ramsey or echo protocol. This limited c_S to a generic coherence comparison and blocked a phase-sensitive phi_S interpretation.

These limitations were not treated as minor metadata omissions. They directly changed the admissible interpretation. A smooth decay curve could not supply seconds, and familiar T2 notation could not supply a pulse sequence.

The targeted qubit-3 test

The primary object-level replay was then restricted to the Fig. 1, qubit 3 lane from Zenodo 11034817. Four linked notebook objects were retained:

Rabi-amplitude calibration

T1 relaxation

Ramsey coherence

Echo coherence

The source notebooks reported approximately

T1 = 245.6 ± 84.0 microseconds

T2R = 89.4 ± 28.7 microseconds

T2E = 242.2 ± 88.5 microseconds

The principal observation is the protocol contrast: the reported Ramsey coherence time is substantially shorter than the echo coherence time. Within standard decoherence analysis, this is consistent with a phase-sensitive contribution that is partly refocused by the echo sequence. It does not identify a unique microscopic noise mechanism, and it does not constitute a new decay law.

For the bridge, the same contrast prevented c_S and phi_S from collapsing into one quantity. The echo surface could constrain a carrier-coherence question, while the Ramsey–echo difference could constrain a phase-sensitive question. T1 remained a separate population-retention surface.

The development package also included predeclared manifest-level controls. Shuffling identity labels changed 263 of 452 indexed relations. Removing the readout shell reduced the shell-bearing rows from 131 to zero. Scrambling the drift order changed 207 of 209 temporal relations. Under these controls, the intended response was abstention, not reinterpretation through another coordinate.

The result is limited by the available package. The notebook-level values and object relations were replayable, but the slim package did not contain the full primary raw trace containers required for an independent trace-level physical pass or fail. The route therefore supports an object-level separation and abstention result. It does not support physical validation of the bridge.

How the data calibrated the Sense Model

The qubit analysis constrained the model in five specific ways.

q_S became protocol-bound. It could be compared with T1 only when preparation, delay, readout, fit form, unit, and qubit identity were specified. A decay-shaped trace alone was insufficient.

c_S became protocol-sensitive. Ramsey, echo, and generic T2 surfaces could no longer be treated as equivalent instances of one coherence coordinate.

phi_S became conditional. It could be compared with Ramsey–echo separation or with a valid Tphi-type diagnostic only after relaxation, units, and protocol compatibility were established.

mu_S became an admission condition for state interpretation. Rabi calibration, resonator response, I/Q semantics, reference classes, and classifier validity had to remain visible. Readout quality could not be averaged into a coherence quantity.

tau_S divided into several temporal objects. Dynamic evolution within a pulse sequence, acquisition order, wall-clock time, and calibration stability were no longer represented by one unqualified time-like coordinate.

This is the metric attachment obtained from the experiment. The physical quantities did not become Sense variables. They restricted the operations that the Sense variables were allowed to support.

The analysis also clarified the remaining specificity problem. A careful experimental physicist already distinguishes T1, Ramsey, echo, readout, and drift. The current result therefore does not establish a unique physical ontology for the Sense Model. Its narrower contribution is a common non-absorption and abstention rule: a missing shell cannot be repaired by a coherence value, a missing protocol cannot be repaired by notation, and a missing time unit cannot be repaired by curve shape.

Current status

The qubit route remains a frozen structural bridge.

The source-linked notebook replay supports the separation of population relaxation, protocol-dependent coherence, phase-sensitive loss, readout, and temporal context. The controls support abstention when identity, shell, or ordering information is removed.

The route does not establish new qubit physics, a device model, a quantum-error-correction threshold, or a physical identification of q_S, c_S, phi_S, mu_S, and tau_S.

A stronger test requires a complete raw dataset from one device and calibration context, a fixed analysis parser declared before inspection, explicit units and protocols, readout calibration, uncertainty propagation, and an independent baseline that performs the same decomposition without the Sense grammar. The bridge becomes physically distinctive only if it adds a reproducible constraint, blocker, or residual structure that the generic analysis does not provide.

The value of the qubit case is therefore not that it makes the Sense Model appear quantum. It is that it makes the model answer to protocol-bound experimental distinctions.