You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/examples/notebooks/simulation/leakage-aware_decoding.ipynb
+3-3Lines changed: 3 additions & 3 deletions
Original file line number
Diff line number
Diff line change
@@ -36,7 +36,7 @@
36
36
"## Introduction\n",
37
37
"The [Local Clustering Decoder](https://arxiv.org/pdf/2411.10343) is the first surface code decoder that retains the performance advantage offered by hardware decoders, while obtaining levels of accuracy and flexibility competitive with software counterparts. LCD is an adaptive and distributed version of an error-clustering algorithm based on Union-Find. Adaptivity refers to updating LCD's error prior based on runtime events, such as heralded leaked qubit measurements. In this notebook, we demonstrate the benefits of LCD by decoding a surface code patch under a circuit-level noise model with leakage — a damaging correlated noise channel affecting most qubit types. We observe a significant improvement in the error-correction performance when leakage adaptivity is included, effectively halving the code distance d required for computation.\n",
38
38
"\n",
39
-
"Leakage is a problematic source of noise in quantum circuits. It refers to the tendency for qubits to drift out of the computational states ($|0\\rangle$ and $|1\\rangle$) to higher energy leaked states ($|2\\rangle$, $|3\\rangle$, etc). Unlike erasure, leakage is silent until measurement. Leakage also differs from conventional Pauli noise — it is long-lived and may spread subsequent noise to other qubits through multi-qubit interactions. This leads to correlated errors in spacetime, which can severely degrade the exponential suppression of logical error that comes with increasing numbers of data qubits in quantum error correction (QEC). Therefore, it is imperative to mitigate the damage of leakage noise both through circuit design with leakage-reduction units and designing decoders that consider its effects. See for example [Coping with qubit leakage in topological codes](https://journals.aps.org/pra/abstract/10.1103/PhysRevA.88.042308)."
39
+
"Leakage is a problematic source of noise in quantum circuits. It refers to the tendency for qubits to drift out of the computational states ($|0\\rangle$ and $|1\\rangle$) to higher energy leaked states ($|2\\rangle$, $|3\\rangle$, etc). Unlike erasure, leakage is silent until measurement. Leakage also differs from conventional Pauli noise — it is long-lived and may spread subsequent noise to other qubits through multi-qubit interactions. This leads to correlated errors in spacetime, which can severely degrade the exponential suppression of logical error that comes with increasing numbers of data qubits in quantum error correction (QEC). Therefore, it is imperative to mitigate the damage of leakage noise both through circuit design with leakage-reduction units and designing decoders that consider its effects. See for example [Coping with qubit leakage in topological codes](https://arxiv.org/abs/1308.6642)."
40
40
]
41
41
},
42
42
{
@@ -67,7 +67,7 @@
67
67
"## Read-in circuits\n",
68
68
"We read in circuits for quantum memory experiments in a $d \\times d \\times d$ surface code, where $d \\in \\{5, 7, 9, 11, 13\\}$. These circuits are subject to a noise model with leakage. Our noise model is an extension of the SI1000 noise model, as described in [Local Clustering Decoder: a fast and adaptive hardware decoder for the surface code](https://arxiv.org/abs/2411.10343). SI1000 is a one-parameter model describing noise in superconducting systems. This parameter $p$ can be taken to be the noise of 2-qubit gates. Our noise model supplements SI1000 with a second parameter, $p_l$, representing the probability that a 2-qubit gate causes a qubit to leak. Leakage may also occur upon reset. A leaked qubit fully depolarises other qubits via 2-qubit gates. Once a qubit is leaked, it can relax back to the computational subspace through relaxation processes, which have their own probability. Eg, qubit resets relax leaked qubits with 100% probability, while 2-qubit gates relax with the same probability they leak: $p_l$.\n",
69
69
"\n",
70
-
"Our circuits contain a leakage reduction unit (LRU) known as patch wiggling, introduced in [Relaxing Hardware Requirements for Surface Code Circuits using Time-dynamics](https://quantum-journal.org/papers/q-2023-11-07-1172/#). This LRU is a small modification in the standard syndrome-extraction (SE) circuits that achieves a role-exchange between data and auxiliary qubits at every round SE. This ensures that every physical qubit is reset every two rounds of SE, flushing regularly leakage from all qubits, and limiting the longevity of leakage to two rounds of SE.\n",
70
+
"Our circuits contain a leakage reduction unit (LRU) known as patch wiggling, introduced in [Relaxing Hardware Requirements for Surface Code Circuits using Time-dynamics](https://arxiv.org/abs/2302.02192). This LRU is a small modification in the standard syndrome-extraction (SE) circuits that achieves a role-exchange between data and auxiliary qubits at every round SE. This ensures that every physical qubit is reset every two rounds of SE, flushing regularly leakage from all qubits, and limiting the longevity of leakage to two rounds of SE.\n",
71
71
"\n",
72
72
"We explore two different parametrisations of our SI1000+leakage noise model, where leakage is both a dominant and subdominant error mechanism by setting $p$ and $p_l$ as follows:\n",
"In our circuits, leakage and relaxation processes are represented as noise channels, that our leakage simulator can interpret. Our simulator is an extension of [Stim: a fast stabilizer circuit simulator](https://quantum-journal.org/papers/q-2021-07-06-497/#). The figure below is an example circuit diagram for a 3x3 surface code with wiggling; run for 2 rounds. In the diagram, the\n",
81
+
"In our circuits, leakage and relaxation processes are represented as noise channels, that our leakage simulator can interpret. Our simulator is an extension of [Stim: a fast stabilizer circuit simulator](https://arxiv.org/abs/2103.02202). The figure below is an example circuit diagram for a 3x3 surface code with wiggling; run for 2 rounds. In the diagram, the\n",
Copy file name to clipboardExpand all lines: docs/guide/authentication.md
+2-2Lines changed: 2 additions & 2 deletions
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -93,10 +93,10 @@ jobs:
93
93
If the token is not set, or not provided as an environment variable, usage of the cloud API will raise an exception:
94
94
95
95
```text
96
-
RuntimeError: Token could not be found neither in ([...]/deltakit-explorer/.env) nor environment variable (DELTAKIT_TOKEN). Please obtain your token at https://deltakit.rivelane.com/dashboard/token and use `Client.set_token` function to register it.
96
+
RuntimeError: Token could not be found neither in ([...]/deltakit-explorer/.env) nor environment variable (DELTAKIT_TOKEN). Please obtain your token at https://deltakit.riverlane.com/dashboard/token and use `Client.set_token` function to register it.
97
97
```
98
98
99
-
If you see this message, please visit the [Deltakit token generation page](https://deltakit.rivelane.com/dashboard/token), generate a token, and set it with `Client.set_token(...)`.
99
+
If you see this message, please visit the [Deltakit token generation page](https://deltakit.riverlane.com/dashboard/token), generate a token, and set it with `Client.set_token(...)`.
0 commit comments