REVIEW 2 major objections 6 minor 3 cited by
Postselection-free experimental observation of the measurement-induced phase transition in circuits with universal gates
T0 review · 2 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash
Pith's one-line read This paper reports a postselection-free experimental observation of the measurement-induced phase transition in circuits with universal gates, using tree-shaped circuits and a linear-cost decoder.
desk verdict A clean proof-of-principle for postselection-free MIPT on a tree, with honest small-size data and a real but unquantified noise caveat in the estimator's unbiasedness. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central object is the dynamical quantum tree: a binary-tree tensor network in which each node applies a two-qubit entangling circuit (a CNOT gate flanked by four Haar-random single-qubit unitaries) followed by weak measurements of tunable strength $\theta$ on both output qubits. The recursion uses the fact that any node's output state is computed from two independent sub-tree outputs, and near criticality only the linear part $Z_t = A_1 Z'_{t-1} + A_2 Z''_{t-1} + O(Z^2)$ matters; the statistics of the coefficients $A_1,A_2$ determine a wavefront speed $v_\theta$ whose zero defines the critical point via $\mathbb{E}[A_1+A_2]=1$. A second load-bearing element is the collapse-process decoder: reversing the flow of time and feeding states back through the tree reconstructs the probe Bloch vector with cost linear in the qubit number, which is what makes the postselection-free estimator possible.
What would settle it
Run the same protocol at fixed $t$ and $\theta$ while injecting known decoherence into the circuits, and compare the empirical distribution of reconstructed Bloch-vector directions conditioned on their magnitudes against the uniform-on-sphere prediction; a significant correlation, or a systematic offset between $\hat{Z}_{t,N_c}(\theta)$ and a brute-force classical evaluation of $Z_t(\theta)$ for the same small $t$, would falsify the unbiasedness premise.
Extended reading notes
Core claim
On the paper's own terms, the central discovery is that the MIPT in a dynamical quantum tree can be detected through the order parameter $Z_t(\theta) = \mathbb{E}_{U,M_w}[Z_t(\theta,U,M_w)]$, the ensemble-averaged smallest eigenvalue of the probe qubit's reduced state, rather than through the typical value $Z_t^{\mathrm{typ}}(\theta)$ that requires postselection. The key identity is $Z_t(\theta) = \mathbb{E}[\tfrac12 - (-1)^{m_0}/\operatorname{sign}(n_t^z)]$, so the order parameter can be estimated by comparing one measured root-qubit outcome $m_0$ with a classically reconstructed Bloch vector $\vec{n}_t$ obtained by running the tree in the collapse direction. Because the tree is recursive, the reconstruction costs $O(N)$ and never requires the same measurement record twice. For this ensemble of weak measurements with Kraus operators $K_m = \sin(\theta/2)\,I + [\cos(\theta/2)-\sin(\theta/2)]\,|m\rangle\langle m|$, the recursion for the order parameter linearizes near the transition and yields the exact critical point $\theta_c = 2.2142(2)$ and the critical scaling $\ln Z_t^{\mathrm{typ}} \sim -t^{1/3}$. The experimental data for $t \le 4$ follow the simulated $Z_t(\theta)$ curves and approach the $t\to\infty$ asymptotic curve, which the paper presents as closing the gap between analytical theory and postselection-free experimental observation of the MIPT with universal gates.
Load-bearing premise
The estimate is unbiased only if the direction of the reconstructed Bloch vector is statistically independent of its length and uniformly distributed on the sphere, as the paper assumes follows from Haar invariance; if device noise or finite-size correlations break that independence, the experimental order parameter would be a biased estimator of $Z_t(\theta)$.
Editorial extensions
If this is right
- The MIPT order parameter can be measured for circuits with universal, Haar-random gates without exponentially many circuit repetitions or an exponential classical simulation, so experiments can scale with the available number of qubits.
- The tree geometry provides an exactly solvable critical point $\theta_c \approx 2.2142$ and a predicted universality class, giving a precision benchmark for quantum hardware.
- Experimental data on a trapped-ion machine for trees up to four layers match theory within 95% confidence intervals without error mitigation, suggesting the protocol is robust to realistic device noise.
- Because the estimator averages linear quantities, a single measurement record per circuit suffices, and data for smaller trees can be recycled by truncating larger measurement records.
- The phase transition can be framed as a predictability transition, predicting the root qubit's initial state from the mid-circuit weak-measurement record, so the probe qubit is not actually required in the experiment.
Reading between the lines
- A natural extension not performed in the paper is to test the independence of Bloch-vector direction and magnitude directly by comparing the empirical joint distribution of $(|\vec{n}_t|, \hat{n}_t)$ from the collapse decoder with the uniform-on-sphere assumption; a detectable correlation would signal a bias in the estimator.
- The same recursive decoding scheme should extend to other recursive tensor geometries and to quantum trees with Abelian or non-Abelian symmetries, where sharpening transitions are predicted to accompany the MIPT.
- A sharper falsification would come from running the protocol on larger trees (for example $t=5$ or $t=6$) and checking whether the crossing point moves toward $\theta_c$ at the predicted rate and whether $\ln Z_t^{\mathrm{typ}}$ obeys the predicted $-t^{1/3}$ scaling at criticality.
- The linear-cost decoder suggests a practical device-noise diagnostic: agreement between hardware results and pool-method simulations of the same circuit ensemble could be monitored as a function of noise levels without any error mitigation.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper constructs a binary-tree quantum circuit model (expansion and collapse processes) built from Haar-random single-qubit gates and weak measurements of tunable strength, and proposes a postselection-free estimator \hat Z_{t,N_c}(θ) (Algorithm 1) for the average minimal eigenvalue Z_t(θ) of the probe qubit. Using rotational invariance of the Haar ensemble, Eq. (6) proves the estimator is unbiased, and the classical decoder is linear in the number of qubits. The authors derive an exact critical point θ_c = 2.2142(2) via a Fisher-KPP-type mapping, simulate finite-t order parameters with the pool method, and compare the simulated curves with data from the Quantinuum H1-1 for t ≤ 4, N_c = 834, and N_s = 8, reporting good agreement without error mitigation.
Significance. If the results hold, the protocol is a significant advance: it bypasses the exponential postselection overhead for monitored circuits with universal (Haar-random) gates on tree geometries, while providing an exactly solvable critical point and critical scaling. The estimator proof in Eq. (6) is clean and self-contained, and the theoretical curves are parameter-free in the sense that no fitting to the experimental data is performed. The public data and code repository is a further strength. Two caveats should be stated fairly: the experiment is limited to t ≤ 4, so it demonstrates finite-size signatures rather than an asymptotic phase transition, and the agreement validates the hardware implementation of the model rather than providing an independent test of the MIPT. These caveats are partially acknowledged in the text and do not by themselves invalidate the protocol claim.
major comments (2)
- [Section V.A and Eq. (6)] The proof that \hat Z_{t,N_c}(θ) is unbiased relies on the conditional Born-rule identity E_{m0|U,Mw}[(-1)^{m0}] = n_z(θ,U,Mw), and on the Haar-invariance argument that makes the Bloch-vector magnitude and direction independent. On the real device, m0 is sampled from a noisy circuit while n_t is computed from the ideal dynamics, so this identity is not exact; the phrase "assuming the noise in the Quantinuum H1-1 is not too large" is not a substitute for a quantitative statement. Because the central claim of the paper is that the experimental data are "precisely described by theory," I ask for an explicit error-model or systematic-error analysis: for example, simulate Algorithm 1 with depolarizing, dephasing, and readout noise at the device's calibrated parameters and show that the induced bias in \hat Z is below the reported 1.96 E_{t,N_c} error bars, or provide a hard bound on the bias derived from measured gate and measurement fidelities.
- [Section III and Fig. 5] The agreement in Fig. 5 is assessed only visually, and no goodness-of-fit statistic is reported. The t ≤ 3 curves are obtained by truncating the same t = 4 measurement records (Section V.D), so the four datasets are statistically correlated, and the text does not explain how this correlation affects the comparison. A quantitative comparison, e.g. a reduced chi-square or maximum deviation expressed in units of the reported 95% confidence intervals, with a statement about how truncation correlations are handled, would be needed to support the claim that the experimental results are "precisely described by theory."
minor comments (6)
- [Algorithm 1, line 9] The quantity sign[n_z(θ,U,Mw)] is undefined when n_z = 0; please state the convention used (e.g., probability-zero event in the ideal Haar ensemble) or specify a numerical cutoff for the experimental analysis.
- [Fig. 5 caption] The caption should explicitly distinguish the dashed curves for Z_t(θ) from the dot-dashed curve for Z_{t→∞}(θ); the current legend text is easy to misread because both curve styles appear in the same figure.
- [Section V.D] The HQC accounting is confusing: the text says the 994 circuits consumed approximately 13099 HQCs while the allocation was 11000 HQCs, and then states that 834 circuits were run within that allocation; please clarify how the remaining 160 circuits were funded and how the total HQC estimate is reconciled with the allocation.
- [Eq. (17)] The numerical evaluation of the expectation over U and M used to obtain θ_c = 2.2142(2) should be described (e.g., Monte Carlo sample size, quadrature, or pool method), and the source of the uncertainty in the last digit should be stated, since this is presented as an exact theoretical result.
- [Section IV] The phrase "expectation values of any function of the probe qubit density matrix are invariant to any single-qubit rotation" is imprecise; it should read "invariant under the Haar average over single-qubit rotations," since a fixed rotation changes the density matrix unless the average is taken.
- [Section V.B] There is a typo in "efficiently (as as function of N)"; it should read "efficiently (as a function of N)."
Circularity Check
No circular derivation: the estimator identity is proved from Haar invariance and the critical point is computed from model statistics, with no parameter fitted to the experimental data.
full rationale
The central estimator result, Eq. (6), is a self-contained derivation: Algorithm 1's X_t is shown to satisfy E[X_t]=Z_t(θ) using the Born-rule identity (7) and the Haar-invariance argument that the Bloch-vector direction is independent of its magnitude. The quantity Z_t(θ) is defined independently in Eq. (4) from the model ensemble, so the estimator is constructed to match a pre-existing target, not fitted to the data. The reported critical point θc=2.2142(2) is obtained by numerically solving Eq. (17) from the statistics of A1 and A2 over the Haar ensemble and Born-rule outcomes; no experimental data point enters this calculation. The comparison in Fig. 5 is therefore a benchmark of the Quantinuum H1-1 implementation against the model's own prediction, which is a standard experimental validation rather than a circular reduction. The paper does rely on the authors' prior work Refs. [16,25] for the linearization theorem and the value λ=1 (Section V.F), and it explicitly concedes that the coefficient statistics are extracted 'despite ignoring the effects of the inputs'; this is a provenance and correctness caveat about the analytic theory, not a circular step, since the cited results are published, assumed, and not redefined in terms of the experimental output. Likewise, the unquantified assumption 'assuming the noise in the Quantinuum H1-1 is not too large' (Section V.A) and the absent proof of pool-method efficiency (Section V.G) are limitations, not circularities. No equation in the paper reduces a prediction to a fitted parameter or to the measured order parameter by construction.
Assumptions & free parameters
assumptions (6)
- domain assumption Only the linear terms of the Z recursion relation (Eq. 14) determine the critical point and scaling exponents.
- domain assumption The coefficients A1 and A2 have identical statistics across all nodes even though measurement statistics depend on the input states.
- domain assumption The velocity-stability condition gives λ=1 at the critical point for the Born-rule sampled collapse process.
- domain assumption The pool method with a finite pool of Z values converges to the true distribution and gives unbiased estimates of Z_t(θ).
- domain assumption The Quantinuum H1-1 noise is small enough that the measured order parameter is not significantly biased.
- domain assumption The CNOT+four-Haar-single-qubit entangling circuit belongs to the same MIPT universality class as a two-qubit Haar-random entangling gate.
Cite this review
Pith. "Pith review of Postselection-free experimental observation of the measurement-induced phase transition in circuits with universal gates." pith.science (2026). https://pith.science/paper/E7LF7RVN
@misc{pith2026250201735,
author = {Pith},
title = {Pith review of: Postselection-free experimental observation of the measurement-induced phase transition in circuits with universal gates},
year = {2026},
howpublished = {\url{https://pith.science/paper/E7LF7RVN}},
note = {Machine review of arXiv:2502.01735}
}
read the original abstract
Monitored many-body systems can exhibit a phase transition between entangling and disentangling dynamical phases by tuning the strength of measurements made on the system as it evolves. This phenomenon is called the measurement-induced phase transition (MIPT). Understanding the properties of the MIPT is a prominent challenge for both theory and experiment at the intersection of many-body physics and quantum information. Realizing the MIPT experimentally is particularly challenging due to the postselection problem, which demands a number of experimental realizations that grows exponentially with the number of measurements made during the dynamics. Proposed approaches that circumvent the postselection problem typically rely on a classical decoding process that infers the final state based on the measurement record. But the complexity of this classical process generally also grows exponentially with the system size unless the dynamics is restricted to a fine-tuned set of unitary operators. In this work we overcome these difficulties. We construct a tree-shaped quantum circuit whose nodes are Haar-random unitary operators followed by weak measurements of tunable strength. For these circuits, we show that the MIPT can be detected without postselection using only a simple classical decoding process whose complexity grows linearly with the number of qubits. Our protocol exploits the recursive structure of tree circuits, which also enables a complete theoretical description of the MIPT, including an exact solution for its critical point and scaling behavior. We experimentally realize the MIPT on Quantinuum's H1-1 trapped-ion quantum computer and show that the experimental results are precisely described by theory. Our results close the gap between analytical theory and postselection-free experimental observation of the MIPT.
Figures
Figures from the paper (4 more)
Forward citations
Cited by 3 Pith papers
-
Anomalous waiting-time distributions in postselection-free quantum many-body dynamics under continuous monitoring
The half-chain waiting-time distribution in a continuously monitored Heisenberg chain develops a non-Poissonian tail whose decay rate is given by the dominant eigenvalue of the no-jump superoperator L0.
-
The Quantum Ensemble Variational Optimization Algorithm: Applications to Molecular Inverse Design
A variational quantum algorithm is reported to search molecular space for drug candidates, with simulations pointing to anticancer molecules using a shallow, few-qubit circuit.
-
Noisy Monitored Quantum Circuits
A review showing that in noisy monitored quantum circuits, any noise enforces area-law entanglement with characteristic q^{-1/3} scaling and noise-correlation-dependent information-protection timescales.
Reference graph
Works this paper leans on
-
[1]
Skinner, J
B. Skinner, J. Ruhman, and A. Nahum, Phys. Rev. X 9, 031009 (2019)
2019
-
[2]
Y. Li, X. Chen, and M. P. A. Fisher, Phys. Rev. B 98, 205136 (2018)
2018
-
[3]
A. Chan, R. M. Nandkishore, M. Pretko, and G. Smith, Phys. Rev. B 99, 224307 (2019)
2019
-
[4]
Y. Li, X. Chen, and M. P. A. Fisher, Phys. Rev. B 100, 134306 (2019)
2019
-
[5]
S. Choi, Y. Bao, X.-L. Qi, and E. Altman, Phys. Rev. Lett. 125, 030505 (2020)
2020
-
[6]
M. J. Gullans and D. A. Huse, Phys. Rev. X 10, 041020 (2020)
2020
-
[7]
Y. Bao, S. Choi, and E. Altman, Phys. Rev. B 101, 104301 (2020)
2020
-
[8]
Jian, Y.-Z
C.-M. Jian, Y.-Z. You, R. Vasseur, and A. W. W. Ludwig, Phys. Rev. B 101, 104302 (2020)
2020
Show all 61 references
-
[9]
Y. Li, X. Chen, A. W. W. Ludwig, and M. P. A. Fisher, Phys. Rev. B 104, 104305 (2021)
2021
-
[10]
Zabalo, M
A. Zabalo, M. J. Gullans, J. H. Wilson, S. Gopalakrish- nan, D. A. Huse, and J. H. Pixley, Phys. Rev. B 101, 060301 (2020)
2020
-
[11]
M. J. Gullans and D. A. Huse, Phys. Rev. Lett. 125, 070606 (2020)
2020
-
[12]
Tang and W
Q. Tang and W. Zhu, Phys. Rev. Research 2, 013022 (2020)
2020
-
[13]
Turkeshi, R
X. Turkeshi, R. Fazio, and M. Dalmonte, Phys. Rev. B 102, 014315 (2020)
2020
-
[14]
R. Fan, S. Vijay, A. Vishwanath, and Y.-Z. You, Phys. Rev. B 103, 174309 (2021)
2021
-
[15]
Li and M
Y. Li and M. P. A. Fisher, Phys. Rev. B 103, 104306 (2021)
2021
-
[16]
Nahum, S
A. Nahum, S. Roy, B. Skinner, and J. Ruhman, PRX Quantum 2, 010352 (2021)
2021
-
[17]
Y. Li, R. Vasseur, M. P. A. Fisher, and A. W. W. Ludwig, Phys. Rev. B 109, 174307 (2024)
2024
-
[18]
Lavasani, Y
A. Lavasani, Y. Alavirad, and M. Barkeshli, Nat. Phys. 17, 342 (2021)
2021
-
[19]
Ippoliti, M
M. Ippoliti, M. J. Gullans, S. Gopalakrishnan, D. A. Huse, and V. Khemani, Phys. Rev. X 11, 011030 (2021)
2021
-
[20]
Lopez-Piqueres, B
J. Lopez-Piqueres, B. Ware, and R. Vasseur, Phys. Rev. B 102, 064202 (2020). 15
2020
-
[21]
C. Noel, P. Niroula, D. Zhu, A. Risinger, L. Egan, D. Biswas, M. Cetina, A. V. Gorshkov, M. J. Gullans, D. A. Huse, and C. Monroe, Nat. Phys. 18, 760 (2022)
2022
-
[22]
Ippoliti and V
M. Ippoliti and V. Khemani, Phys. Rev. Lett. 126, 060501 (2021)
2021
-
[23]
Y. Li, Y. Zou, P. Glorioso, E. Altman, and M. P. A. Fisher, Phys. Rev. Lett. 130, 220404 (2023)
2023
-
[24]
Y. Li, S. Vijay, and M. P. Fisher, PRX Quantum 4, 010331 (2023)
2023
-
[25]
X. Feng, B. Skinner, and A. Nahum, PRX Quantum 4, 030333 (2023)
2023
-
[26]
Google Quantum AI and Collaborators, Nature 622, 481 (2023)
2023
-
[27]
M. Fava, L. Piroli, T. Swann, D. Bernard, and A. Nahum, Phys. Rev. X 13, 041045 (2023)
2023
-
[28]
C.-M. Jian, H. Shapourian, B. Bauer, and A. W. W. Lud- wig, Measurement-induced entanglement transitions in quantum circuits of non-interacting fermions: Born-rule versus forced measurements (2023), arXiv:2302.09094
2023 arXiv
-
[29]
Agrawal, A
U. Agrawal, A. Zabalo, K. Chen, J. H. Wilson, A. C. Potter, J. H. Pixley, S. Gopalakrishnan, and R. Vasseur, Phys. Rev. X 12, 041002 (2022)
2022
-
[30]
Barratt, U
F. Barratt, U. Agrawal, A. C. Potter, S. Gopalakrishnan, and R. Vasseur, Phys. Rev. Lett. 129, 200602 (2022)
2022
-
[31]
Majidy, U
S. Majidy, U. Agrawal, S. Gopalakrishnan, A. C. Potter, R. Vasseur, and N. Y. Halpern, Phys. Rev. B108, 054307 (2023)
2023
-
[32]
J. M. Koh, S.-N. Sun, M. Motta, and A. J. Minnich, Nat. Phys. 19, 1314 (2023)
2023
-
[33]
Ippoliti and V
M. Ippoliti and V. Khemani, PRX Quantum 5, 020304 (2024)
2024
-
[34]
A. A. Akhtar, H.-Y. Hu, and Y.-Z. You, Phys. Rev. B 109, 094209 (2024)
2024
-
[35]
S. J. Garratt and E. Altman, PRX Quantum 5, 030311 (2024)
2024
-
[36]
McGinley, PRX Quantum 5, 020347 (2024)
M. McGinley, PRX Quantum 5, 020347 (2024)
2024
-
[37]
Agrawal, J
U. Agrawal, J. Lopez-Piqueres, R. Vasseur, S. Gopalakr- ishnan, and A. C. Potter, Phys. Rev. X 14, 041012 (2024)
2024
-
[38]
Kamakari, J
H. Kamakari, J. Sun, Y. Li, J. J. Thio, T. P. Gujarati, M. P. A. Fisher, M. Motta, and A. J. Minnich, Exper- imental demonstration of scalable cross-entropy bench- marking to detect measurement-induced phase transi- tions on a superconducting quantum processor (2024), arXiv:2403.00938
2024 arXiv
-
[39]
A. C. Potter and R. Vasseur, Entanglement dynamics in hybrid quantum circuits, in Entanglement in Spin Chains: From Theory to Quantum Technology Applica- tions, edited by A. Bayat, S. Bose, and H. Johannes- son (Springer International Publishing, Cham, 2022) pp. 211–249
2022
-
[40]
M. P. Fisher, V. Khemani, A. Nahum, and S. Vijay, An- nual Review of Condensed Matter Physics14, 335 (2023)
2023
-
[41]
Fidkowski, J
L. Fidkowski, J. Haah, and M. B. Hastings, Quantum 5, 382 (2021)
2021
-
[42]
Szyniszewski, A
M. Szyniszewski, A. Romito, and H. Schomerus, Phys. Rev. B 100, 064204 (2019)
2019
-
[43]
Szyniszewski, A
M. Szyniszewski, A. Romito, and H. Schomerus, Phys. Rev. Lett. 125, 210602 (2020)
2020
-
[44]
Cˆ ot´ e and S
J. Cˆ ot´ e and S. Kourtis, Nat Commun13, 7395 (2022)
2022
-
[45]
X. Feng, N. Fishchenko, S. Gopalakrishnan, and M. Ip- politi, Charge and spin sharpening transitions on dynam- ical quantum trees (2024), arXiv:2405.13894
2024 arXiv
-
[46]
G. M. Sommers, D. A. Huse, and M. J. Gullans, Dy- namically generated concatenated codes and their phase diagrams (2024), arXiv:2409.13801
2024
-
[47]
J. D. Miller and B. Derrida, J Stat Phys 75, 357 (1994)
1994
-
[48]
Monthus and T
C. Monthus and T. Garel, J. Phys. A: Math. Theor. 42, 075002 (2009)
2009
-
[49]
Garc ´ ıa-Mata, O
I. Garc ´ ıa-Mata, O. Giraud, B. Georgeot, J. Martin, R. Dubertrand, and G. Lemari´ e, Phys. Rev. Lett. 118, 166801 (2017)
2017
-
[50]
Shi, L.-M
Y.-Y. Shi, L.-M. Duan, and G. Vidal, Phys. Rev. A 74, 022320 (2006)
2006
-
[51]
V. Murg, F. Verstraete, O. Legeza, and R. M. Noack, Phys. Rev. B 82, 205105 (2010)
2010
-
[52]
Silvi, V
P. Silvi, V. Giovannetti, S. Montangero, M. Rizzi, J. I. Cirac, and R. Fazio, Phys. Rev. A 81, 062335 (2010)
2010
-
[53]
Tagliacozzo, G
L. Tagliacozzo, G. Evenbly, and G. Vidal, Phys. Rev. B 80, 235127 (2009)
2009
-
[54]
W. Li, J. von Delft, and T. Xiang, Phys. Rev. B 86, 195137 (2012)
2012
-
[55]
Quantinuum H1-1, https://www.quantinuum.com/ (2024)
2024
-
[56]
Derrida and D
B. Derrida and D. Simon, EPL 78, 60006 (2007)
2007
-
[57]
Derrida and H
B. Derrida and H. Spohn, J Stat Phys 51, 817 (1988)
1988
-
[58]
Banaszek, M
K. Banaszek, M. Cramer, and D. Gross, New J. Phys. 15, 125020 (2013)
2013
-
[59]
Sivarajah, S
S. Sivarajah, S. Dilkes, A. Cowtan, W. Simmons, A. Edg- ington, and R. Duncan, Quantum Sci. Technol.6, 014003 (2020)
2020
-
[60]
X. Feng, J. Cˆ ot´ e, S. Kourtis, and B. Skinner, https://doi.org/10.5281/zenodo.14767528 (2025), Zen- odo dataset
2025 doi
-
[106]
Given t, data points are ( θ, Zt(θ)) for 50 different mea- surement strengths θ
Increasing values of t correspond to darker blue. Given t, data points are ( θ, Zt(θ)) for 50 different mea- surement strengths θ. For clarity, we omit the markers for the data points and only show the connecting lines. We do include error bars for each data point correspond- ...
Reviewed August 9, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.