Quantum Trojan Insertion: Controlled Activation for Covert Circuit Manipulation

Jayden John; Lakshman Golla; Qian Wang

REVIEW 4 major objections 5 minor 2 cited by

Quantum Trojan Insertion: Controlled Activation for Covert Circuit Manipulation

T0 review · 4 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper proposes a controllable quantum Trojan that an untrusted compiler can embed in empty circuit slots, staying dormant until triggered by a control signal.

desk verdict The paper's central claim of a conditionally activated quantum Trojan is not implemented in its own algorithm; every inserted gate is unconditional, though the empirical observation about gate insertion is real and worth repurposing. read the letter →

arxiv 2502.08880 v1 pith:ACPHRSW6 submitted 2025-02-13 quant-ph cs.AR

Jayden John , Lakshman Golla , Qian Wang This is my paper · ORCID

classification quant-phcs.AR MSC 81P68

keywords quantumtrojanhardwarecircuitcompilationcontrolledactivationsecurityemptyslotinsertiondepth-preservingattack

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that an untrusted quantum compiler can plant a Trojan that behaves like a classical hardware Trojan: the malicious gates stay inactive during normal operation and only alter the computation when a switch qubit is set. The insertion routine finds idle slots in the circuit's layered structure, places an X gate on a control qubit in the first column as the switch, and fills later empty positions with controlled-NOT (CX) gates as the payload, so the circuit's depth is unchanged. On the benchmark circuits used, the authors report the altered output distributions differ from the originals by a total variation distance close to 90%, with only about 20% more gates. If this is right, a user checking the circuit on standard inputs would see correct outputs while a triggered run silently computes something else.

What carries the argument

Algorithm 1 is the carrier of the argument: it converts the circuit to a directed acyclic graph, extracts temporal layers, computes empty positions as the set difference between all qubits and the qubits used in each layer, then inserts an X gate on a designated control qubit in column 0 and random CX gates from that control to empty target qubits in later columns until a gate limit is reached. The X gate is the switch that is supposed to give the Trojan conditional activation, and the CX gates are the payload that flips target qubits only when the control qubit is in |1⟩. The empty-slot discipline is what keeps the circuit depth unchanged.

What would settle it

Run the Trojan-inserted circuit with the control qubit in the |0> state and compute the total variation distance between its output distribution and that of the original circuit; if that distance is not near zero, the switch does not actually deactivate the Trojan.

Watch

Extended reading notes

Core claim

The central claim is that a quantum Trojan can be made controllable and dormant: an X gate placed on a control qubit at column 0 acts as the trigger switch, and CX gates placed in later empty columns act as the payload, so the circuit computes exactly the original function while the switch is off and a different function when it is on. The paper presents this as a compiler-side attack: a DAG-based scan of the circuit finds empty slots, and Trojan gates are inserted only there, preserving the critical-path depth. The authors position this as the first step beyond static single-gate Trojans, which are easy to remove during optimization or to flag with machine-learning detectors.

Load-bearing premise

The load-bearing premise is that an X gate on the control qubit at the first column works as a switch, letting the same compiled circuit run either dormant (control qubit in |0>) or active, even though the paper gives no mechanism for choosing which mode runs and a fixed circuit applies every gate on every run.

Editorial extensions

If this is right

A malicious compiler could add 2-6 gates per benchmark circuit without increasing circuit depth.
Existing static-gate detection methods and compiler optimization passes, which remove redundant gates, are less likely to catch a payload that is meant to be latent.
Circuits with more qubits and deeper structure provide more empty slots, which is where the reported total variation distance approaches 1.
The threat model requires only compiler access, matching the standard remote-quantum-service setup.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A reader can test the dormancy claim directly by running the inserted circuit with the control qubit initialized to |0> and comparing its output distribution with the original; the paper reports only the altered (switch-on) runs, so this check is the natural next experiment.
A stronger attack would tie the trigger to an actual input qubit rather than a dedicated switch, so the activation condition depends on the user's data and leaves no extra wires to audit.
Because the switch qubit appears to be an extra wire at the top of the circuit, a full-stack verifier could look for added wires as a signature, a defense the paper does not discuss.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The reader's take is right, and the stress-test note pins down the load-bearing flaw exactly. The paper promises a quantum Trojan that stays dormant until triggered by specific input conditions, but Algorithm 1 inserts an unconditional X gate at column zero and CX gates in later empty slots. In standard quantum circuit semantics, every gate applies on every execution, so there is no off state and no trigger. The 'Accuracy Deactivated' column in Table I makes it worse: those deviations from the original (-4% to +5.7%) show that removing the switch does not restore the original function. The paper's own text in Section II-D and IV asserts conditional activation, but the method contradicts it.

What is actually new? The specific recipe of a leading X 'switch' plus CX gates in empty DAG slots is not verbatim in the cited prior work, which used static X or SWAP insertions. The paper also correctly notes that placing gates in empty slots avoids depth overhead; that observation is real but is a direct consequence of the insertion rule, not an independent finding.

Where the paper does get credit: the TVD measurements are direct simulations of circuits the authors constructed, so they are internally consistent. The empirical fact that inserting a few CX gates into empty slots can substantially alter output distributions on noisy hardware is credible and could be reused in a better-designed study. The literature review is adequate and the citation pattern seems honest.

The soft spots are not minor. The central claim of controllable activation is unsupported by the experiments, there is no detection experiment despite the paper claiming resilience to detection, and no code or data were shipped. The 0% depth increase is enforced by the insertion algorithm, so it is not a meaningful overhead result. The only genuinely useful piece is the empirical demonstration that empty slots are a viable insertion point, which is a modest building block, not a novel Trojan design.

This paper is for researchers in quantum compiler security who might want to see an example of how not to implement a conditional trigger, or who might reuse the empty-slot observation in a more rigorous study. It deserves a serious referee because the question it targets is important and the empirical data could inform future work, but as submitted it should be rejected: the core claim is internally inconsistent with the algorithm. If you review it, insist that the authors either provide a concrete mechanism for conditional activation (e.g., measurement-based or classically conditioned gates) or drop the conditional language entirely and reframe the contribution as static gate insertion into empty slots.

Referee Report

4 major / 5 minor

Summary. The paper proposes a method for a malicious quantum compiler to insert 'controlled Trojans' into quantum circuits. Algorithm 1 converts a circuit to a DAG, finds empty slots in each temporal layer, inserts an X gate on a control qubit at column 0, and then inserts random CX gates in later empty columns up to a gate limit. The authors claim these Trojans remain dormant until triggered by predefined input conditions, can be activated or deactivated, add zero circuit depth, and cause large total variation distance (TVD) from the original circuit. The experiments simulate several RevLib benchmarks with Qiskit's FakeValencia noise model and report depth, gate count, accuracy, and TVD values in Table I and Figure 5.

Significance. If a genuinely dormant, conditionally triggered quantum Trojan with zero depth overhead existed, it would be an important security concern for untrusted compilation. The paper also has some strengths: the threat model is clearly stated, the DAG-based empty-slot insertion strategy is concrete, and the use of standard benchmarks with a noise model gives the experimental section a reasonable shape. However, the central claim of controlled, conditional activation is not realized by the proposed construction. Algorithm 1 produces circuits in which every gate executes on every run; there is no input-dependent or runtime-trigger mechanism. The significance of the paper therefore rests on an unsupported and, as written, internally inconsistent premise. The reported 'deactivated' behavior appears to be a different circuit rather than a dormant mode of the same compiled circuit.

major comments (4)

[Section IV-A, Algorithm 1] Algorithm 1 inserts an X gate at column 0 and CX gates at subsequent columns with no conditional logic. In standard quantum circuit execution every gate applies on every run; there is no measurement-based feedback, classical control flow, or input-dependent predicate that could skip the X gate or the CX gates. Since all qubits are initialized to |0>, the X gate flips the control qubit to |1> and the subsequent CX gates act unconditionally. Thus the circuit produced by Algorithm 1 is always in the 'activated' state, and the claim that the Trojan remains dormant until triggered is not implemented.
[Section IV, Figure 3 and Table I] The 'Trojan Switch OFF' behavior shown in Figure 3 is not a mode of the same compiled circuit; it appears to be a circuit in which the X gate is omitted. No experiment demonstrates a single circuit that behaves as the original under one condition and as the altered circuit under another. Table I's 'Accuracy Deactivated' column is not accompanied by an explanation of how deactivation is realized or verified; if it corresponds to removing the X gate, the residual accuracy changes of -4% to +5.7% need to be accounted for (e.g., as sampling or backend noise) before this column can support the restoration claim.
[Section II-D, IV-B, V-C] The paper never specifies what the 'predefined input conditions' or 'specific control signal conditions' are. The authors need to exhibit a concrete input or control-signal setting that leaves the output distribution unchanged and another setting that triggers the Trojan, using the same compiled circuit in both cases. Without such a demonstration, the asserted analogy to classical hardware Trojans with rare-input triggers is unsupported. A minimal test would be to use the control qubit as a data-dependent input or to implement the switch with mid-circuit measurement and feed-forward, and then compare output distributions under the two settings.
[Section V-D, Table I] The '0% depth increase' result is a direct consequence of the insertion rule (gates are only placed in previously empty DAG slots), not a measured effect of an otherwise invasive modification. This is not itself a flaw, but the paper presents it as evidence that the Trojan is stealthy while the more relevant security property—conditional dormancy within one compiled circuit—is absent. The depth claim should be reframed as a design property of the insertion strategy rather than as evidence for the controlled-activation contribution.

minor comments (5)

[Algorithm 1] The pseudocode uses 'available qubits' before it is initialized; Step 2 begins with 'available qubits ← available qubits ∩ empty positions' but no initial value for 'available qubits' is given. This should be fixed (e.g., initialize to all qubits at the start of each column).
[Section V-B, Eq. (1)] The sentence says TVD is 'normalized by the total number of shots' but the formula has denominator 2N, which is the correct normalization for total variation distance. The wording should be corrected to avoid confusion with an L1 distance normalized by N.
[Section V-C and Figure 5] The text and Figure 5 use the term 'restored circuit' without defining it, while Table I uses 'Accuracy Deactivated.' The manuscript should use consistent terminology and explain how each 'restored' or 'deactivated' circuit is constructed.
[Section II-B] There are several grammar and wording issues, e.g., 'a series of quantum gates (functional units) and that manipulate,' and 'operator as U, which transform the states.' These should be edited for clarity.
[Section V-A] The paper does not state whether the code, circuit construction scripts, or random seeds are available. Since the insertion involves random target selection and the table reports averages over 20 iterations, reproducibility would be improved by providing this information.

Circularity Check

1 steps flagged · score 2.0 of 10

The 0% depth-increase result is guaranteed by the empty-slot insertion rule rather than independently measured, but the central Trojan claims are not circular; they are unsupported correctness assertions.

self definitional [Section V-D (Cost and Overhead Analysis); Algorithm 1, Step 2]
"we implemented a selection algorithm that strategically places Trojan gates exclusively in unused or empty slots within the circuit. This approach ensures that the circuit depth remains unchanged after Trojan insertion, preserving its functional timing and critical path integrity."

The 0% depth-increase claim is not an independent empirical result; it is a logical consequence of the insertion rule. Algorithm 1 defines empty positions as qubits with no operation in a given DAG layer and inserts gates only into those positions. Adding a gate to an empty slot cannot lengthen the critical path, so the depth is forced to remain identical before any simulation is run. The paper presents this construction tautology as an experimental outcome under 'Cost and Overhead Analysis,' even though the depth equality is already guaranteed by the definition of 'empty slots.' This does not affect the TVD measurements, which are direct simulations, but it does make the overhead claim self-definitional.

full rationale

The paper's central claim—controllable, conditionally dormant Trojans—is not established by a derivation: Algorithm 1 inserts X and CX gates unconditionally, and standard quantum circuit semantics apply every gate on every run. That is a correctness and feasibility gap rather than a circularity, because no quantity is fitted to data and later renamed as a prediction, and no load-bearing premise is supported only by the authors' own prior citations. The one result that does reduce by construction is the 0% depth-increase claim: since the insertion routine places gates only in empty slots, the depth is identical before measurement, and reporting this as a finding is a self-definitional overhead claim rather than an independent discovery. The TVD and accuracy values are direct simulations of circuits the authors constructed, so those are measurements, not circular predictions. Overall, there is minor definitional circularity in the overhead analysis, but the main security claims are not circular in the technical sense used here.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new physical entities; the 'Trojan control gate' is an ordinary X gate. The central unstated assumptions are about switch semantics and the control qubit state, and they carry the conditional-activation claim.

free parameters (2)

gate_limit = 2 to 6 inserted gates per circuit (Table I)
The number of inserted X and CX gates is an input to Algorithm 1 chosen per benchmark; it determines the magnitude of the output distribution shift and is not derived from any stated principle.
control_position = 0 (qubit index 0)
Algorithm 1 takes 'control pos' as an input and the paper fixes it to qubit 0 without justification; this choice determines whether and how the CX gates affect the computation.

assumptions (4)

domain assumption The control qubit used for the switch is initialized to |0> at circuit start, so the inserted X gate deterministically flips it to |1>.
Invoked by Figure 3 and Algorithm 1; the paper never states this explicitly, and if the control qubit is a data input, the CX gates are not uniformly dormant.
ad hoc to paper An X gate placed at column 0 acts as a runtime switch that can be 'off' even though the gate is present in the compiled circuit.
This is the central enabling assumption of the conditional-activation claim in Section IV-A; in standard quantum circuit execution, every gate in the circuit applies on every run.
domain assumption The untrusted compiler has access to both the original and transpiled circuits and can add gates during compilation.
Threat model stated in Section III; this aligns with prior work [5], [11] but is not independently verified for real quantum cloud pipelines.
ad hoc to paper Filling empty DAG slots with CX gates leaves the circuit functionally unchanged when the switch is off.
Used by Algorithm 1 to claim dormancy; false in general because CX gates change the state whenever their control qubit is in |1>.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Quantum Trojan Insertion: Controlled Activation for Covert Circuit Manipulation." pith.science (2026). https://pith.science/paper/ACPHRSW6

@misc{pith2026250208880,
  author       = {Pith},
  title        = {Pith review of: Quantum Trojan Insertion: Controlled Activation for Covert Circuit Manipulation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/ACPHRSW6}},
  note         = {Machine review of arXiv:2502.08880}
}

read the original abstract

Quantum computing has demonstrated superior efficiency compared to classical computing. Quantum circuits are essential for implementing functions and achieving correct computational outcomes. Quantum circuit compilers, which translate high-level quantum operations into hardware-specific gates while optimizing performance, serve as the interface between the quantum software stack and physical quantum machines. However, untrusted compilers can introduce malicious hardware Trojans into quantum circuits, altering their functionality and leading to incorrect results. In the world of classical computing, effective hardware Trojans are a critical threat to integrated circuits. This process often involves stealthily inserting conditional logic gates that activate under specific input conditions. In this paper, we propose a novel advanced quantum Trojan that is controllable, allowing it to be activated or deactivated under different circumstances. These Trojans remain dormant until triggered by predefined input conditions, making detection challenging. Through a series of benchmark experiments, we demonstrate the feasibility of this method by evaluating the effectiveness of embedding controlled trojans in quantum circuits and measuring their impact on circuit performance and security.

Figures

Figures reproduced from arXiv: 2502.08880 by the authors.

**Figure 1.** An example of a quantum circuit C. Quantum Circuit Compilation and Trojans Quantum circuit compilation transforms a circuit representing a quantum algorithm into a hardware-executable format, adapting it to the quantum hardware topology, connectivity, and error characteristics of the target quantum hardware. This process is analogous to compiling classical programs but must address unique quantum constraints such a… view at source ↗

**Figure 2.** Illustration of the untrusted compiler threat model. During the compile process, the malicious Trojans are inserted. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Example of Trojan insertion. When the switch is off (left), [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 5.** Figure 5: Distribution of Total Variation Distance (TVD) of benchmark [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

SQUASH: A SWAP-Based Quantum Attack to Sabotage Hybrid Quantum Neural Networks
quant-ph 2025-06 reject novelty 4.0 of 10

Inserting SWAP gates into the variational circuit of a hybrid quantum neural network degrades classification accuracy by up to roughly 74%, with targeted insertions able to ruin a single class's accuracy.
Quantum Properties Trojans (QuPTs) for Attacking Quantum Neural Networks
quant-ph 2025-07 reject novelty 2.0 of 10

Quantum Properties Trojans insert extra unitary or Hadamard gates into a QNN circuit, causing accuracy to drop (up to 23% relative) and, for the Hadamard attack, forcing the model to predict a single class.

Reference graph

Works this paper leans on

28 extracted references · 24 canonical work pages · cited by 2 Pith papers

[1]

Ibm quantum breaks the 100-qubit processor barrier,

J. Chow, O. Dial, and J. Gambetta, “Ibm quantum breaks the 100-qubit processor barrier,” IBM Research Blog , vol. 2, 2021

work page 2021
[2]

Cloud based qc with amazon braket,

C. Gonzalez, “Cloud based qc with amazon braket,” Digitale Welt, vol. 5, no. 2, pp. 14–17, 2021

work page 2021
[3]

Quantum programming on azure quantum—an open source tool for quantum developers,

K. Prateek and S. Maity, “Quantum programming on azure quantum—an open source tool for quantum developers,” in Quantum Computing: A Shift from Bits to Qubits . Springer, 2023, pp. 283–309

work page 2023
[4]

Trojannet: Detecting trojans in quantum circuits using machine learning,

S. Das and S. Ghosh, “Trojannet: Detecting trojans in quantum circuits using machine learning,” arXiv preprint arXiv:2306.16701 , 2023

arXiv 2023
[5]

Hardware trojans in quantum circuits, their impacts, and defense,

R. Roy, S. Das, and S. Ghosh, “Hardware trojans in quantum circuits, their impacts, and defense,” in 2024 25th International Symposium on Quality Electronic Design (ISQED) . IEEE, 2024, pp. 1–8

work page 2024
[6]

Multi-Stage Watermarking for Quantum Circuits

M. Yang, X. Guo, and L. Jiang, “Multi-stage watermarking for quantum circuits,” arXiv preprint arXiv:2404.18038 , 2024

work page Pith review arXiv 2024
[7]

Exploration of power side-channel vulnerabilities in quantum computer controllers,

C. Xu, F. Erata, and J. Szefer, “Exploration of power side-channel vulnerabilities in quantum computer controllers,” in Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security, 2023, pp. 579–593

work page 2023
[8]

Dynamic pulse switching for protection of quantum computation on untrusted clouds,

T. Trochatos, S. Deshpande, C. Xu, Y . Lu, Y . Ding, and J. Szefer, “Dynamic pulse switching for protection of quantum computation on untrusted clouds,” in 2024 IEEE International Symposium on Hardware Oriented Security and Trust (HOST) . IEEE, 2024, pp. 404–414

work page 2024

Show all 28 references

[9]

A primer on security of quantum computing,

S. Ghosh, S. Upadhyay, and A. A. Saki, “A primer on security of quantum computing,” arXiv preprint arXiv:2305.02505 , 2023

2023 arXiv
[10]

Secure quantum circuit compilation methodology for untrusted compilers,

S. Das and S. Ghosh, “Secure quantum circuit compilation methodology for untrusted compilers,” 2024

2024
[11]

Stealthy swaps: Adversarial swap injection in multi-tenant quantum computing,

S. Upadhyay and S. Ghosh, “Stealthy swaps: Adversarial swap injection in multi-tenant quantum computing,” in 2024 37th International Con- ference on VLSI Design and 2024 23rd International Conference on Embedded Systems (VLSID) . IEEE, 2024, pp. 474–479

2024
[12]

Trojan attacks on variational quantum circuits and countermeasures,

S. Das and S. Ghosh, “Trojan attacks on variational quantum circuits and countermeasures,” in 2024 25th International Symposium on Quality Electronic Design (ISQED) , 2024, pp. 1–8

2024
[13]

Trustworthy hardware: Identifying and classifying hardware trojans,

R. Karri, J. Rajendran, K. Rosenfeld, and M. Tehranipoor, “Trustworthy hardware: Identifying and classifying hardware trojans,” IEEE Com- puter, vol. 43, no. 10, pp. 39–46, 2010

2010
[14]

Bhunia and M

S. Bhunia and M. Tehranipoor, Hardware Security: A Hands-on Learn- ing Approach. Morgan Kaufmann, 2014

2014
[15]

A comprehensive solution for hardware trojan defense,

D. Forte, A. Srivastava, and S. Bhunia, “A comprehensive solution for hardware trojan defense,” Proceedings of the IEEE , vol. 105, no. 3, pp. 284–311, 2017

2017
[16]

Hardware trojan detection using path delay fin- gerprint,

Y . Jin and Y . Makris, “Hardware trojan detection using path delay fin- gerprint,” in 2008 IEEE International Workshop on Hardware-Oriented Security and Trust , 2008, pp. 51–57

2008
[17]

A2: Analog malicious hardware,

K. Yang, M. Hicks, Q. Dong, T. Austin, and D. Sylvester, “A2: Analog malicious hardware,” in 2016 IEEE Symposium on Security and Privacy (SP), 2016, pp. 18–37

2016
[18]

Security against hardware trojan through a novel application of design obfuscation,

R. S. Chakraborty and S. Bhunia, “Security against hardware trojan through a novel application of design obfuscation,” in Proceedings of the 2009 International Conference on Computer-Aided Design , 2009, pp. 113–116

2009
[19]

Sarlock: Sat attack resistant logic locking,

M. Yasin, B. Mazumdar, O. Sinanoglu, and D. Forte, “Sarlock: Sat attack resistant logic locking,” IEEE Transactions on VLSI Systems , vol. 25, no. 10, pp. 2930–2943, 2017

2017
[20]

Case study: Detecting hardware trojans in third-party digital ip cores,

J. Zhang, M. Tehranipoor, and D. Forte, “Case study: Detecting hardware trojans in third-party digital ip cores,” in 2014 IEEE International Symposium on Hardware-Oriented Security and Trust (HOST) , 2014, pp. 56–59

2014
[21]

Mapping the patent landscape of quantum technologies: Patenting trends, innovation and policy implica- tions,

M. Aboy, T. Minssen, and M. Kop, “Mapping the patent landscape of quantum technologies: Patenting trends, innovation and policy implica- tions,” IIC-International Review of Intellectual Property and Competi- tion Law, vol. 53, no. 6, pp. 853–882, 2022

2022
[22]

Randomized reversible gate-based obfus- cation for secured compilation of quantum circuit,

S. Das and S. Ghosh, “Randomized reversible gate-based obfus- cation for secured compilation of quantum circuit,” arXiv preprint arXiv:2305.01133, 2023

2023 arXiv
[23]

Short paper: A quantum circuit obfuscation methodology for security and privacy,

A. Suresh, A. A. Saki, M. Alam, R. Onur Topaloglu, and S. Ghosh, “Short paper: A quantum circuit obfuscation methodology for security and privacy,” in Proceedings of the 10th International Workshop on Hardware and Architectural Support for Security and Privacy , 2021, pp. 1–5

2021
[24]

Quantum logic locking for security,

R. O. Topaloglu, “Quantum logic locking for security,” MDPI Multidis- ciplinary Scientific Journal , vol. 6, no. 3, pp. 411–420, 2023

2023
[25]

An open- source, industrial-strength optimizing compiler for quantum programs,

R. S. Smith, E. C. Peterson, M. G. Skilbeck, and E. J. Davis, “An open- source, industrial-strength optimizing compiler for quantum programs,” Quantum Science and Technology , vol. 5, no. 4, p. 044001, 2020

2020
[26]

Automating the comparison of quantum compilers for quantum circuits,

M. Salm, J. Barzen, F. Leymann, B. Weder, and K. Wild, “Automating the comparison of quantum compilers for quantum circuits,” in Sympo- sium and Summer School on Service-Oriented Computing . Springer, 2021, pp. 64–80

2021
[27]

Revlib: An online resource for reversible functions and reversible circuits,

R. Wille, D. Große, L. Teuber, G. W. Dueck, and R. Drechsler, “Revlib: An online resource for reversible functions and reversible circuits,” in 38th International Symposium on Multiple Valued Logic (ismvl 2008) . IEEE, 2008, pp. 220–225

2008
[28]

Quantum computing with Qiskit,

A. Javadi-Abhari, M. Treinish, K. Krsulich, C. J. Wood, J. Lishman, J. Gacon, S. Martiel, P. D. Nation, L. S. Bishop, A. W. Cross, B. R. Johnson, and J. M. Gambetta, “Quantum computing with Qiskit,” 2024

2024

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

Ibm quantum breaks the 100-qubit processor barrier,

J. Chow, O. Dial, and J. Gambetta, “Ibm quantum breaks the 100-qubit processor barrier,” IBM Research Blog , vol. 2, 2021

work page 2021

[2] [2]

Cloud based qc with amazon braket,

C. Gonzalez, “Cloud based qc with amazon braket,” Digitale Welt, vol. 5, no. 2, pp. 14–17, 2021

work page 2021

[3] [3]

Quantum programming on azure quantum—an open source tool for quantum developers,

K. Prateek and S. Maity, “Quantum programming on azure quantum—an open source tool for quantum developers,” in Quantum Computing: A Shift from Bits to Qubits . Springer, 2023, pp. 283–309

work page 2023

[4] [4]

Trojannet: Detecting trojans in quantum circuits using machine learning,

S. Das and S. Ghosh, “Trojannet: Detecting trojans in quantum circuits using machine learning,” arXiv preprint arXiv:2306.16701 , 2023

arXiv 2023

[5] [5]

Hardware trojans in quantum circuits, their impacts, and defense,

R. Roy, S. Das, and S. Ghosh, “Hardware trojans in quantum circuits, their impacts, and defense,” in 2024 25th International Symposium on Quality Electronic Design (ISQED) . IEEE, 2024, pp. 1–8

work page 2024

[6] [6]

Multi-Stage Watermarking for Quantum Circuits

M. Yang, X. Guo, and L. Jiang, “Multi-stage watermarking for quantum circuits,” arXiv preprint arXiv:2404.18038 , 2024

work page Pith review arXiv 2024

[7] [7]

Exploration of power side-channel vulnerabilities in quantum computer controllers,

C. Xu, F. Erata, and J. Szefer, “Exploration of power side-channel vulnerabilities in quantum computer controllers,” in Proceedings of the 2023 ACM SIGSAC Conference on Computer and Communications Security, 2023, pp. 579–593

work page 2023

[8] [8]

Dynamic pulse switching for protection of quantum computation on untrusted clouds,

T. Trochatos, S. Deshpande, C. Xu, Y . Lu, Y . Ding, and J. Szefer, “Dynamic pulse switching for protection of quantum computation on untrusted clouds,” in 2024 IEEE International Symposium on Hardware Oriented Security and Trust (HOST) . IEEE, 2024, pp. 404–414

work page 2024