Error Mitigation for Near-Term Devices
Extracting better answers from noisy hardware without full QEC.
Without full error correction, a noisy quantum computer returns biased answers. Error mitigation does not repair the quantum state; it repairs the statistics, by running extra or modified circuits and combining the results into a better estimate. The bill is paid in measurement shots, and it grows exponentially with circuit size.
What you’ll learn
- Distinguish error mitigation (better estimates of expectation values) from error correction (protected quantum states).
- Correct readout errors with a confusion matrix and quantify the variance cost.
- Apply zero-noise extrapolation and probabilistic error cancellation, and compute their sampling overheads.
- Use symmetry verification to flag errors, and explain which errors it misses.
- Explain why every mitigation method pays a cost exponential in the total noise of the circuit.
Estimate, don't correct
Most near-term (NISQ) algorithms, such as VQE, need expectation values rather than individual bit strings. Noise replaces the ideal state by a noisy one, and the estimate becomes biased: no number of shots averages the bias away.
Error mitigation builds an estimator from many runs of noisy (possibly modified) circuits such that . It uses no extra qubits for encoding and no real-time feedback. The error of any estimator splits as
| Method | Idea | Needs |
|---|---|---|
| Readout mitigation | Invert the measured confusion matrix | Calibration circuits |
| Zero-noise extrapolation (ZNE) | Measure at amplified noise, extrapolate to zero | A way to scale the noise |
| Probabilistic error cancellation (PEC) | Sample from a signed decomposition of the inverse noise | An accurate noise model |
| Symmetry verification | Discard shots that violate a known conserved quantity | A symmetry of the ideal state |
Readout errors
Measurement is often the noisiest operation. For one qubit, let and . Measured frequencies are a linear image of the true ones, , with the confusion matrix
Substitute and . The signal shrinks by and shifts.
Inverting is unbiased (if is accurate), but it divides the statistical fluctuations by too.
For a parity with symmetric errors the factors multiply: the overhead is exponential in the number of qubits measured.
A qubit has symmetric readout error . You measure . What is the mitigated estimate?
Show the answer
C.
. The price: the shot noise is divided by as well, so you need times the shots for the same error bar.
Zero-noise extrapolation
Suppose you could turn the noise up by a known factor . Measuring at several values of and fitting a curve lets you extrapolate to , the noiseless value. Ways to amplify noise:
- Pulse stretching: run every gate more slowly with proportionally weaker pulses, so decoherence acts times longer (the gates must be recalibrated).
- Gate folding: replace a gate by , logically the same but with three times the noise; folding only some gates gives intermediate .
- Probabilistic error amplification: insert extra random Pauli errors drawn from a learned noise model.
The extrapolation is only as good as its model. Noise often makes signals decay exponentially in the amount of noise, and a straight line through exponential data undershoots. You can watch such a decay directly: in the experiment below the coherence of decays like (dephasing), and stretching a circuit's duration by multiplies the exponent by .
A signal decays exponentially with the noise scale: you measure and , and the true value is . What does linear extrapolation, , return?
Show the answer
0.84
: much better than the raw , but still biased. Exponential extrapolation, , is exact here because the model matches the decay.
Probabilistic error cancellation
If you know the noise channel that follows a gate, you can try to undo it. is usually not a physical operation, but it can be written as a signed combination of operations you can implement, typically Pauli gates:
Insert after the gate and expand by linearity.
For each shot, pick with probability , run the circuit with inserted, and record times the measured outcome. The average is unbiased.
The recorded values are times larger, so the variance, and with it the shot count, grows by about .
For single-qubit depolarizing noise with Pauli error probability per gate, (derived in the Rydberg State lesson). Over gates the costs multiply.
| Total noise | PEC overhead |
|---|---|
Each of 500 gates has . Roughly how many times more shots does PEC need?
- About
- About
- About
- About
Show the answer
A. About
, and the shot overhead is .
Symmetry verification, and the limits
Sometimes you know a property the ideal state must have: a conserved particle number, a fixed parity. Measure it, and throw away the shots that violate it. The Bell state , for instance, has even parity: .
Symmetry verification removes the errors that break the symmetry, at the price of discarded shots. The surviving fraction decays exponentially with the number of error locations, so this method, too, has an exponential cost.
Key takeaways
- Error mitigation estimates expectation values with reduced bias from many noisy runs; it does not produce protected states.
- Readout mitigation inverts a confusion matrix; ZNE extrapolates to ; PEC samples a quasi-probability decomposition of ; symmetry verification postselects on conserved quantities.
- Each removes bias at a variance cost: for readout, for Richardson extrapolation, for PEC.
- All costs grow exponentially with the expected number of errors per run, so mitigation suits circuits with errors; large computations need error correction.
Check yourself
In the interactive lesson these are scored with your confidence, and what you miss comes back for review.
What does error mitigation deliver?
- A corrected quantum state that can be used in further computation
- Logical qubits with lower error rates
- Exact noiseless results at no extra cost
- Better estimates of expectation values, assembled from many runs of noisy circuits
Show the answer
D. Better estimates of expectation values, assembled from many runs of noisy circuits
Mitigation post-processes measurement statistics. It never produces a cleaner quantum state, so it cannot feed a longer computation the way error correction does.
At noise scales and you measure and . What does linear extrapolation to give?
Show the answer
B.
The slope is per unit of , so . Equivalently, .
Why is the cost of probabilistic error cancellation exponential in circuit size?
- The quasi-probability norms of the gates multiply, , and the shot count scales as
- Each gate must be characterized by tomography with exponentially many measurements
- The number of possible error patterns grows exponentially, and PEC enumerates all of them
- PEC needs one extra qubit per gate
Show the answer
A. The quasi-probability norms of the gates multiply, , and the shot count scales as
Each estimate is multiplied by , so its variance grows like . Sampling avoids enumerating error patterns, but not this blow-up.
Symmetry verification using the parity of a Bell state catches which single-qubit errors?
- All single-qubit errors
- Only errors
- and errors, but not errors
- None, because measuring the parity destroys the entanglement
Show the answer
C. and errors, but not errors
An error is caught when it anticommutes with : and on either qubit do, while commutes and slips through. Measuring only the parity leaves the Bell state intact.
Go deeper
- Zero-Noise Extrapolation & PEC · Amplify noise to remove it; invert it with quasi-probabilities.
- The Limits of Error Mitigation · Exponential sampling costs and what they imply.
Further reading
- K. Temme, S. Bravyi & J. M. Gambetta, "Error mitigation for short-depth quantum circuits", Phys. Rev. Lett. 119, 180509 (2017).
- A. Kandala, K. Temme, A. D. Córcoles, A. Mezzacapo, J. M. Chow & J. M. Gambetta, "Error mitigation extends the computational reach of a noisy quantum processor", Nature 567, 491 (2019).
- Z. Cai, R. Babbush, S. C. Benjamin, S. Endo, W. J. Huggins, Y. Li, J. R. McClean & T. E. O'Brien, "Quantum error mitigation", Rev. Mod. Phys. 95, 045005 (2023).
- Y. Kim et al., "Evidence for the utility of quantum computing before fault tolerance", Nature 618, 500 (2023).