pith. sign in

arxiv: 2605.25868 · v1 · pith:A4HJI2WCnew · submitted 2026-05-25 · 💻 cs.HC · cs.LG

The Timing Dependencies of Trust: Speed, Accuracy, and cBCI Neuro-Decoupling in Human-AI Teams

Christopher Baker , Stephen Hinton , Akashdeep Nijjar , Riccardo Poli , Caterina Cinel , Tom Reed , Stephen Fairclough This is my paper

Pith reviewed 2026-06-29 20:34 UTC · model grok-4.3

classification 💻 cs.HC cs.LG
keywords Human-AI teamscBCItrust dynamicsneuro-decouplingRiemannian OracleVirtual Reality drone taskAI timingcollaborative interfaces
0
0 comments X

The pith

AI timing dictates whether human teams enter blind compliance or delayed hesitation in cBCI drone tasks.

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

The paper establishes that the speed of an AI teammate changes the failure states of human-AI teams during high-workload search tasks. Fast but less-accurate AI triggers instant reflexive compliance, collapsing human accuracy to 50.2 percent and limiting pure behavioral teams of size 8 to 74.1 percent. Slow but accurate AI instead produces hesitation and cognitive conflict, with initial accuracy at 61.1 percent yet full recovery to 100 percent in larger behavioral teams. The 2D Adaptive Riemannian Oracle detects these states by narrowing or widening its temporal windows and fuses the isolated signals to lift performance in both conditions.

Core claim

AI timing dictates the mechanism of team failure. Fast AI induced instant, blind compliance; human accuracy under deception collapsed to 50.2 percent, and pure behavioural teams (N=8) failed to scale beyond 74.1 percent. In contrast, Slow AI induced delayed cognitive conflict; humans hesitated (61.1 percent accuracy), but N=8 behavioural teams eventually recovered to 100.0 percent. The Riemannian Oracle mathematically adapted to these states by heavily restricting temporal windows below 0.8 seconds to intercept fast reflexive compliance while widening windows above 1.2 seconds to capture delayed cognitive conflict, enabling hybrid fusion to rescue fast-AI teams by 7.6 percent at N=8 and acce

What carries the argument

The 2D Adaptive Riemannian Oracle, which maps spatial covariance and dynamically restricts temporal windows below 0.8 seconds for fast AI or widens them above 1.2 seconds for slow AI to isolate veridical signals for hybrid fusion.

If this is right

  • Hybrid fusion rescues fast-AI teams by 7.6 percent at team size 8.
  • Smaller slow-AI teams recover faster with a 6.9 percent gain at team size 4.
  • cBCI synergy requires matching oracle temporal windows to the AI speed that produced the failure state.
  • Dynamically gated Human-AI systems become feasible once the oracle intercepts the compliance or conflict states.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Timing could be treated as an explicit design lever for controlling trust and compliance in AI teammates rather than relying solely on accuracy.
  • The same window-adaptation principle might extend to other continuous high-load team settings such as monitoring or remote operation.
  • Varying AI speed and accuracy independently would test whether timing alone drives the observed compliance-versus-hesitation split.

Load-bearing premise

The 2D Adaptive Riemannian Oracle can reliably distinguish and intercept the distinct failure states of reflexive compliance versus delayed cognitive conflict by restricting or widening temporal windows based on AI speed, and that these neuro-decoupled signals improve team performance when fused.

What would settle it

A test in which the oracle-adapted hybrid fusion produces no accuracy gain over baseline behavioral teams in either the fast-AI or slow-AI condition.

read the original abstract

The speed and accuracy of an artificial teammate fundamentally alter the failure states of Human-AI integration. While high-speed AI interventions risk inducing reflexive blind compliance, delayed interventions can induce ambiguous cognitive conflict. This study investigates how the fundamental characteristics of an in-task AI assistant, Fast/Less-Accurate (FLA-AI) versus Slow/Accurate (SA-AI) impact the synergy of Collaborative Brain-Computer Interface (cBCI) teams in a Virtual Reality drone task. Seventeen operators completed continuous search tasks under high cognitive workload while their spatial covariance was mapped using a 2D Adaptive Riemannian Oracle. The results mathematically demonstrate that AI timing dictates the mechanism of team failure. Fast AI induced instant, blind compliance; human accuracy under deception collapsed to 50.2%, and pure behavioural teams (N=8) failed to scale beyond 74.1%. In contrast, Slow AI induced delayed cognitive conflict; humans hesitated (61.1% accuracy), but N=8 behavioural teams eventually recovered to 100.0%. Crucially, the Riemannian Oracle mathematically adapted to these states: it heavily restricted temporal windows (< 0.8s) to intercept fast reflexive compliance, while widening windows (> 1.2s) to capture delayed cognitive conflict. Integrating these isolated veridical signals via Hybrid Fusion successfully rescued the Fast AI team (+7.6% at N=8) and significantly accelerated the recovery of smaller Slow AI teams (+6.9% at N=4). These findings prove that cBCI synergy is heavily contingent on the temporal dynamics of trust, providing a critical framework for designing dynamically gated Human-AI systems.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript reports results from a VR drone search task with 17 operators comparing Fast/Less-Accurate AI (FLA-AI) versus Slow/Accurate AI (SA-AI). It claims that AI timing dictates team failure mechanisms: fast AI produces instant blind compliance (human accuracy 50.2%, N=8 behavioral teams capped at 74.1%), while slow AI produces delayed cognitive conflict (human accuracy 61.1%, N=8 teams recover to 100%). A 2D Adaptive Riemannian Oracle is said to adapt temporal windows (<0.8 s for fast, >1.2 s for slow) to isolate veridical neuro-decoupled signals from spatial covariance, enabling hybrid fusion gains of +7.6% (N=8 fast) and +6.9% (N=4 slow). The central conclusion is that cBCI synergy depends on temporal trust dynamics and that the oracle provides a framework for dynamically gated Human-AI systems.

Significance. If the reported accuracy differences and oracle adaptation hold under proper statistical controls and if the oracle can adapt windows without explicit knowledge of AI speed, the work would supply a concrete, timing-dependent account of trust failure modes in cBCI teams together with a practical gating mechanism. The hybrid-fusion improvements and the mathematical demonstration of window adaptation would be directly usable for designing adaptive human-AI systems. However, the absence of statistical tests, error bars, and participant/exclusion criteria in the reported numbers substantially weakens any such claim at present.

major comments (3)
  1. [Abstract / Results] Abstract and Results sections: the reported accuracies (50.2%, 61.1%, 74.1%, 100.0%) and fusion gains (+7.6%, +6.9%) are presented without statistical tests, confidence intervals, participant counts per condition, or exclusion criteria. These omissions make it impossible to determine whether the claimed differences between fast and slow conditions are reliable or whether they support the assertion that timing dictates distinct failure mechanisms.
  2. [Abstract / Methods (Oracle)] Oracle description (abstract and methods): the 2D Adaptive Riemannian Oracle is stated to 'heavily restrict temporal windows (< 0.8 s) to intercept fast reflexive compliance, while widening windows (> 1.2 s) to capture delayed cognitive conflict.' Because AI speed is an explicit, known experimental factor, this rule-based adjustment does not demonstrate detection of the two cognitive states from spatial covariance features alone. The claim that the oracle isolates 'veridical' neuro-decoupled signals therefore rests on an untested assumption that the same adaptation would occur when AI speed is withheld from the oracle.
  3. [Results (Hybrid Fusion)] Hybrid-fusion results: the reported performance rescues are attributed to integration of 'isolated veridical signals,' yet the manuscript supplies no ablation showing that the gains disappear when the oracle is forced to operate without knowledge of AI speed. This leaves open the possibility that the improvements reflect timing-aware gating rather than cBCI-derived separation of compliance versus conflict states.
minor comments (2)
  1. [Abstract] The abstract states 'mathematically demonstrate' and 'mathematically adapted' without showing any equations, derivations, or parameter definitions; these phrases should be replaced by precise empirical descriptions or the relevant equations should be supplied.
  2. [Figures / Tables] No table or figure caption supplies the exact N per cell, trial counts, or the precise definition of the temporal-window thresholds; these details are required for reproducibility.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive comments, which identify key areas where statistical rigor and methodological transparency can be strengthened. We address each major comment point-by-point below and will revise the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract / Results] Abstract and Results sections: the reported accuracies (50.2%, 61.1%, 74.1%, 100.0%) and fusion gains (+7.6%, +6.9%) are presented without statistical tests, confidence intervals, participant counts per condition, or exclusion criteria. These omissions make it impossible to determine whether the claimed differences between fast and slow conditions are reliable or whether they support the assertion that timing dictates distinct failure mechanisms.

    Authors: We agree that the submitted manuscript omitted statistical tests, confidence intervals, explicit per-condition participant counts, and exclusion criteria. The study enrolled 17 operators total. Behavioral teams used N=8 per condition (FLA-AI and SA-AI), with hybrid fusion evaluated at N=8 (fast) and N=4 (slow). We will add two-sample or paired t-tests (or appropriate non-parametric equivalents), 95% confidence intervals, and a Methods subsection detailing exclusion criteria (EEG signal quality thresholds and minimum task completion) in the revised version. This will allow direct evaluation of the reliability of the reported accuracy differences and failure-mode claims. revision: yes

  2. Referee: [Abstract / Methods (Oracle)] Oracle description (abstract and methods): the 2D Adaptive Riemannian Oracle is stated to 'heavily restrict temporal windows (< 0.8 s) to intercept fast reflexive compliance, while widening windows (> 1.2 s) to capture delayed cognitive conflict.' Because AI speed is an explicit, known experimental factor, this rule-based adjustment does not demonstrate detection of the two cognitive states from spatial covariance features alone. The claim that the oracle isolates 'veridical' neuro-decoupled signals therefore rests on an untested assumption that the same adaptation would occur when AI speed is withheld from the oracle.

    Authors: The referee correctly observes that the window thresholds (<0.8 s for fast, >1.2 s for slow) are set using knowledge of the experimental AI-speed condition. The manuscript presents these adjustments as a mathematical demonstration that the oracle can be tuned to the distinct trust dynamics induced by each timing regime, rather than as an autonomous, blind classifier operating solely on covariance features. We will revise the abstract and Methods to state this limitation explicitly and to frame the oracle as a condition-aware gating mechanism whose parameters are derived from observed behavioral states. We do not claim blind detection in the current results. revision: yes

  3. Referee: [Results (Hybrid Fusion)] Hybrid-fusion results: the reported performance rescues are attributed to integration of 'isolated veridical signals,' yet the manuscript supplies no ablation showing that the gains disappear when the oracle is forced to operate without knowledge of AI speed. This leaves open the possibility that the improvements reflect timing-aware gating rather than cBCI-derived separation of compliance versus conflict states.

    Authors: We agree that the manuscript contains no ablation in which the oracle is run without access to AI-speed information. The reported fusion gains (+7.6% at N=8 fast; +6.9% at N=4 slow) therefore reflect integration under the condition-specific window settings. In revision we will add an explicit limitations paragraph acknowledging that the gains may partly derive from timing-aware gating and will discuss how the oracle framework could be extended to fully unsupervised window adaptation. No post-hoc blind ablation is possible with the existing dataset without new analysis pipelines, but the current results still demonstrate the value of timing-dependent neuro-decoupling within the tested design. revision: partial

Circularity Check

1 steps flagged

Oracle window adaptation rule is constructed from known AI speed input rather than neuro-signal inference.

specific steps
  1. fitted input called prediction [Abstract]
    "Crucially, the Riemannian Oracle mathematically adapted to these states: it heavily restricted temporal windows (< 0.8s) to intercept fast reflexive compliance, while widening windows (> 1.2s) to capture delayed cognitive conflict."

    The window thresholds are set directly from the known experimental condition (fast vs. slow AI), which is an a priori input. The oracle therefore does not infer or distinguish the cognitive states (reflexive compliance vs. delayed conflict) from the Riemannian features alone; the 'isolation of veridical signals' and performance gains are produced by timing-aware gating that is forced by the input factor rather than derived from the neuro-data.

full rationale

The paper presents mostly empirical behavioral results (accuracy percentages under FLA-AI vs SA-AI conditions) that do not reduce to self-referential definitions. However, the load-bearing claim that the 2D Adaptive Riemannian Oracle distinguishes and intercepts distinct failure states via neuro-decoupling reduces to a timing-based rule that takes the experimental AI speed factor as direct input. This matches the fitted_input_called_prediction pattern because the 'adaptation to these states' and resulting hybrid-fusion gains are not independent detections from spatial covariance features. No self-citations, equations, or ansatzes are involved, so the circularity is partial and localized to the oracle mechanism rather than the entire derivation chain.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

Abstract-only review; full methods, equations, and prior citations unavailable. The reported time windows and accuracy thresholds appear to be derived from the experiment rather than pre-specified.

free parameters (1)
  • temporal window thresholds
    The <0.8 s and >1.2 s windows are presented as adapted values that intercept the two failure states; their exact fitting procedure is not described.
axioms (1)
  • domain assumption Spatial covariance patterns measured by the 2D Adaptive Riemannian Oracle can be used to detect and distinguish reflexive compliance from delayed cognitive conflict.
    Invoked to justify the oracle's adaptation to AI timing states.

pith-pipeline@v0.9.1-grok · 5855 in / 1596 out tokens · 32859 ms · 2026-06-29T20:34:32.290388+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

47 extracted references · 7 canonical work pages

  1. [1]

    & Gadiraju, U

    Erlei, A., Sharma, A. & Gadiraju, U. Understanding Choice Independence and Error Types in Human-AI Collaboration. in Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems 1–19 (Association for Computing Machinery, New York, NY, USA, 2024). doi:10.1145/3613904.3641946

    work page doi:10.1145/3613904.3641946 2024

  2. [2]

    & Lindsay, J

    Goldfarb, A. & Lindsay, J. R. Prediction and Judgment: Why Artificial Intelligence Increases the Importance of Humans in War. Int. Secur. 46 , 7–50 (2022)

    2022

  3. [3]

    A., Perer, A

    Cabrera, Á. A., Perer, A. & Hong, J. I. Improving Human-AI Collaboration With Descriptions of AI Behavior. Proc ACM Hum-Comput Interact 7 , 136:1-136:21 (2023)

    2023

  4. [4]

    Li, J. et al. Understanding the Effects of Miscalibrated AI Confidence on User Trust, Reliance, and Decision Efficacy. Preprint at https://doi.org/10.48550/arXiv.2402.07632 (2025)

    work page doi:10.48550/arxiv.2402.07632 2025

  5. [5]

    Roeder, L. et al. A Quantum Model of Trust Calibration in Human–AI Interactions. Entropy 25 , 1362 (2023)

    2023

  6. [6]

    B., Flathmann, C

    Canonico, L. B., Flathmann, C. & McNeese, N. Collectively Intelligent Teams: Integrating Team Cognition, Collective Intelligence, and AI for Future Teaming. Proc. Hum. Factors Ergon. Soc. Annu. Meet. 63 , 1466–1470 (2019)

    2019

  7. [7]

    & Gadiraju, U

    He, G., Buijsman, S. & Gadiraju, U. How Stated Accuracy of an AI System and Analogies to Explain Accuracy Affect Human Reliance on the System. Proc ACM Hum-Comput Interact 7 , 276:1-276:29 (2023)

    2023

  8. [8]

    & Zammit-Mangion, D

    Koopman, C. & Zammit-Mangion, D. Artificial Intelligence for Real-Time Tolerance to Critical Flight Data Errors in Large Aircraft. J. Aerosp. Inf. Syst. 21 , 726–734 (2024)

    2024

  9. [9]

    Possibilities and Challenges for Artificial Intelligence in Military Applications

  10. [10]

    Vox Populi

    GALTON, F. Vox Populi. Nature 75 , 450–451 (1907)

    1907

  11. [11]

    Kerr, N. L. & Tindale, R. S. Group Performance and Decision Making. Annu. Rev. Psychol. 55 , 623–655 (2004)

    2004

  12. [12]

    P., Sienkiewicz, J., Fronczak, A

    Orzechowski, K. P., Sienkiewicz, J., Fronczak, A. & Fronczak, P. When the crowd gets it wrong – the limits of collective wisdom in machine learning. Sci. Rep. 15 , 22139 (2025)

    2025

  13. [13]

    Jayles, B. et al. The impact of incorrect social information on collective wisdom in human groups. J. R. Soc. Interface 17 , 20200496 (2020)

    2020

  14. [14]

    & Schweitzer, F

    Mavrodiev, P. & Schweitzer, F. Enhanced or distorted wisdom of crowds? An agent-based model of opinion formation under social influence. Swarm Intell. 15 , 31–46 (2021)

    2021

  15. [15]

    & Cinel, C

    Poli, R., Valeriani, D. & Cinel, C. Collaborative Brain-Computer Interface for Aiding Decision-Making. PLoS ONE 9 , e102693 (2014)

    2014

  16. [16]

    & Poli, R

    Bhattacharyya, S., Valeriani, D., Cinel, C., Citi, L. & Poli, R. Collaborative Brain-Computer Interfaces to Enhance Group Decisions in an Outpost Surveillance Task. in 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) 3099–3102 (IEEE, Berlin, Germany, 2019). doi:10.1109/EMBC.2019.8856309

    work page doi:10.1109/embc.2019.8856309 2019

  17. [17]

    Are You Really Sure?

    Ma, S. et al. “Are You Really Sure?” Understanding the Effects of Human Self-Confidence Calibration in AI-Assisted Decision Making. in Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems 1–20 (Association for Computing Machinery, New York, NY, USA, 2024). doi:10.1145/3613904.3642671

    work page doi:10.1145/3613904.3642671 2024

  18. [18]

    Z., Zhou, J., Bridon, C., Chen, F

    Arshad, S. Z., Zhou, J., Bridon, C., Chen, F. & Wang, Y. Investigating User Confidence for Uncertainty Presentation in Predictive Decision Making. in Proceedings of the Annual Meeting of the Australian Special Interest Group for Computer Human Interaction 352–360 (Association for Computing Machinery, New York, NY, USA, 2015). doi:10.1145/2838739.2838753

    work page doi:10.1145/2838739.2838753 2015

  19. [19]

    Fleming, S. M. & Lau, H. C. How to measure metacognition. Front. Hum. Neurosci. 8 , (2014)

    2014

  20. [20]

    E., Cleeremans, A

    Wokke, M. E., Cleeremans, A. & Ridderinkhof, K. R. Sure I’m Sure: Prefrontal Oscillations Support Metacognitive Monitoring of Decision Making. J. Neurosci. 37 , 781–789 (2017)

    2017

  21. [21]

    E., Achoui, D

    Wokke, M. E., Achoui, D. & Cleeremans, A. Action information contributes to metacognitive decision-making. Sci. Rep. 10 , 3632 (2020)

    2020

  22. [22]

    B., Kerr, T

    Murayama, K., Blake, A. B., Kerr, T. & Castel, A. D. When enough is not enough: Information overload and metacognitive decisions to stop studying information. J. Exp. Psychol. Learn. Mem. Cogn. 42 , 914–924 (2016)

    2016

  23. [23]

    Di Gregorio, F., Maier, M. E. & Steinhauser, M. Are errors detected before they occur? Early error sensations revealed by metacognitive judgments on the timing of error awareness. Conscious. Cogn. 77 , 102857 (2020)

    2020

  24. [24]

    G., Ahmadipour, P

    Sadras, N., Sani, O. G., Ahmadipour, P. & Shanechi, M. M. Post-stimulus encoding of decision confidence in EEG: toward a brain–computer interface for decision making. J. Neural Eng. 20 , 056012 (2023)

    2023

  25. [25]

    & Gibbons, H

    Schnuerch, R., Schmuck, J. & Gibbons, H. Cortical oscillations and event-related brain potentials during the preparation and execution of deceptive behavior. Psychophysiology 61 , e14695 (2024)

    2024

  26. [26]

    & Summerfield, C

    Yeung, N. & Summerfield, C. Metacognition in human decision-making: confidence and error monitoring. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 367 , 1310–1321 (2012)

    2012

  27. [27]

    & Lamm, C

    Eisma, J., Rawls, E., Long, S., Mach, R. & Lamm, C. Frontal midline theta differentiates separate cognitive control strategies while still generalizing the need for cognitive control. Sci. Rep. 11 , 14641 (2021)

    2021

  28. [28]

    Cohen, M. X. & Donner, T. H. Midfrontal conflict-related theta-band power reflects neural oscillations that predict behavior. J. Neurophysiol. 110 , 2752–2763 (2013)

    2013

  29. [29]

    Cohen, M. X. A neural microcircuit for cognitive conflict detection and signaling. Trends Neurosci. 37 , 480–490 (2014)

    2014

  30. [30]

    F., Martin, L

    Potts, G. F., Martin, L. E., Kamp, S.-M. & Donchin, E. Neural response to action and reward prediction errors: Comparing the error-related negativity to behavioral errors and the feedback-related negativity to reward prediction violations. Psychophysiology 48 , 218–228 (2011)

    2011

  31. [31]

    & Niznikiewicz, M

    Cao, D., Li, Y. & Niznikiewicz, M. A. Neural characteristics of cognitive reappraisal success and failure: An ERP study. Brain Behav. 10 , e01584 (2020)

    2020

  32. [32]

    McDonnell, A. S. et al. This Is Your Brain on Autopilot: Neural Indices of Driver Workload and Engagement During Partial Vehicle Automation. Hum. Factors 65 , 1435–1450 (2023)

    2023

  33. [33]

    A., Bergstedt, D

    Choe, J., Coffman, B. A., Bergstedt, D. T., Ziegler, M. D. & Phillips, M. E. Transcranial Direct Current Stimulation Modulates Neuronal Activity and Learning in Pilot Training. Front. Hum. Neurosci. 10 , (2016)

    2016

  34. [34]

    E., Zaepffel, M., Brovelli, A., MacKay, W

    Kilavik, B. E., Zaepffel, M., Brovelli, A., MacKay, W. A. & Riehle, A. The ups and downs of beta oscillations in sensorimotor cortex. Exp. Neurol. 245 , 15–26 (2013)

    2013

  35. [35]

    & Brown, P

    Tan, H., Wade, C. & Brown, P. Post-Movement Beta Activity in Sensorimotor Cortex Indexes Confidence in the Estimations from Internal Models. J. Neurosci. 36 , 1516–1528 (2016)

    2016

  36. [36]

    Parker, A. et al. Role of uncertainty in sensorimotor control. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 357 , 1137–1145 (2002)

    2002

  37. [37]

    E., Auksztulewicz, R., Ondobaka, S

    Palmer, C. E., Auksztulewicz, R., Ondobaka, S. & Kilner, J. M. Sensorimotor beta power reflects the precision-weighting afforded to sensory prediction errors. NeuroImage 200 , 59–71 (2019)

    2019

  38. [38]

    & Murray, I

    Plainis, S. & Murray, I. J. Reaction times as an index of visual conspicuity when driving at night. Ophthalmic Physiol. Opt. J. Br. Coll. Ophthalmic Opt. Optom. 22 , 409–415 (2002)

    2002

  39. [39]

    Wickens, C. D. Multiple resources and mental workload. Hum. Factors 50 , 449–455 (2008)

    2008

  40. [40]

    Barratt, E. S. Impulsiveness Defined Within a Systems Model of Personality. in Advances in Personality Assessment (Routledge, 1985)

    1985

  41. [41]

    Kothe, C. et al. The Lab Streaming Layer for Synchronized Multimodal Recording. bioRxiv 2024.02.13.580071 (2024) doi:10.1101/2024.02.13.580071

    work page doi:10.1101/2024.02.13.580071 2024

  42. [42]

    Gramfort, A. et al. MEG and EEG data analysis with MNE-python. Front. Neurosci. 7 , 267 (2013)

    2013

  43. [43]

    & de Boer, J

    van Rossum, G. & de Boer, J. Interactively testing remote servers using the Python programming language. CWI Q. 4 , 283–304 (1991)

    1991

  44. [44]

    Data Structures for Statistical Computing in Python

    McKinney, W. Data Structures for Statistical Computing in Python. in 56–61 (Austin, Texas, 2010). doi:10.25080/Majora-92bf1922-00a

    work page doi:10.25080/majora-92bf1922-00a 2010

  45. [45]

    Harris, C. R. et al. Array programming with NumPy. Nature 585 , 357–362 (2020)

    2020

  46. [46]

    Hunter, J. D. Matplotlib: A 2D Graphics Environment. Comput. Sci. Eng. 9 , 90–95 (2007)

    2007

  47. [47]

    Waskom, M. L. seaborn: statistical data visualization. J. Open Source Softw. 6 , 3021 (2021)

    2021