pith. machine review for the scientific record. sign in

arxiv: 2604.23716 · v1 · submitted 2026-04-26 · 💻 cs.AI · cs.IT· cs.LG· cs.MA· math.IT

Recognition: unknown

Information-Theoretic Measures in AI: A Practical Decision Guide

Nikolaos Al.Papadopoulos , Konstantinos E. Psannis
Authors on Pith no claims yet

Pith reviewed 2026-05-08 06:00 UTC · model grok-4.3

classification 💻 cs.AI cs.ITcs.LGcs.MAmath.IT
keywords information-theoretic measuresAI decision frameworkentropymutual informationtransfer entropyintegrated informationestimator selectionagent complexity
0
0 comments X

The pith

A decision framework helps select and apply seven information-theoretic measures in AI by matching each to its question, estimator, and misuse risks.

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

This paper supplies a practical guide for choosing among seven common information-theoretic measures used in artificial intelligence. It organizes the guidance around three questions for each measure: what question it answers in a given AI context, which estimator fits the data type and size, and what misuse poses the greatest risk. The guidance is turned into usable tools: a selection flowchart and a master decision table, plus Bridge Boxes that tie the math to ideas such as uncertainty or directed influence. Three detailed examples show the framework applied to representation learning, temporal influence analysis, and measuring complexity in evolved agents. The overall aim is to make measure selection less arbitrary and more aligned with estimator assumptions and safe claims.

Core claim

The authors supply a decision framework for entropy, cross-entropy, mutual information, transfer entropy, integrated information (Phi), effective information (EI), and autonomy. For each measure the framework states the question it addresses in AI or agent settings, the estimator suited to the data type and dimensionality, and the most dangerous misuse. The framework is delivered through a measure-selection flowchart, a master decision table, and standardized Bridge Boxes that link the mathematical quantities to cognitive constructs. Three worked examples demonstrate its use on representation learning, temporal influence, and agent complexity.

What carries the argument

The three prescriptive questions per measure together with the measure-selection flowchart, master decision table, and Bridge Boxes that link IT quantities to cognitive constructs.

If this is right

  • Estimator choice becomes matched to data type and dimensionality for each measure.
  • The most common misuses, such as treating mutual information as causation, are explicitly flagged and avoided.
  • Both standard machine-learning tasks and decision-making agent domains receive clear guidance.
  • Bridge Boxes produce consistent mapping from quantities like Phi to notions of complexity or autonomy.
  • The flowchart reduces arbitrary selection by narrowing options according to the question being asked.

Where Pith is reading between the lines

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

  • Widespread use could standardize how papers report estimator assumptions when citing these measures.
  • The same structure might be applied to newer information-theoretic quantities that appear in future AI research.
  • The cognitive links in the Bridge Boxes open a route for testing the framework in cognitive-science or neuroscience settings.
  • A follow-up validation study measuring error rates before and after use of the table would test its practical impact.

Load-bearing premise

The seven listed measures are the most relevant ones for AI work and the Bridge Boxes give accurate, non-misleading links between the mathematical quantities and cognitive constructs.

What would settle it

A controlled user study in which practitioners follow the flowchart and table yet still select mismatched estimators or commit the warned-against misuses on standard tasks.

read the original abstract

Information-theoretic (IT) measures are ubiquitous in artificial intelligence: entropy drives decision-tree splits and uncertainty quantification, cross-entropy is the default classification loss, mutual information underpins representation learning and feature selection, and transfer entropy reveals directed influence in dynamical systems. A second, less consolidated family of measures, integrated information (Phi), effective information (EI), and autonomy, has emerged for characterizing agent complexity. Despite wide adoption, measure selection is often decoupled from estimator assumptions, failure modes, and safe inferential claims. This paper provides a practical decision framework for all seven measures, organized around three prescriptive questions for each: (i) what question does the measure answer and in which AI context; (ii) which estimator is appropriate for the data type and dimensionality; and (iii) what is the most dangerous misuse. The framework is operationalized in two complementary artifacts: a measure-selection flowchart and a master decision table. We cover both AI/ML and decision-making agent application domains per measure, with standardized Bridge Boxes linking IT quantities to cognitive constructs. Three worked examples illustrate the framework on concrete practitioner scenarios spanning representation learning, temporal influence analysis, and evolved agent complexity.

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

2 major / 2 minor

Summary. The manuscript proposes a practical decision framework for seven information-theoretic measures commonly used in AI, including entropy, cross-entropy, mutual information, transfer entropy, integrated information (Phi), effective information (EI), and autonomy. The framework is structured around three prescriptive questions for each measure: the question it answers in AI contexts, appropriate estimators for data types, and dangerous misuses. It is operationalized via a measure-selection flowchart, a master decision table, standardized Bridge Boxes linking IT quantities to cognitive constructs, and three worked examples in representation learning, temporal influence analysis, and evolved agent complexity.

Significance. If the synthesis accurately represents the estimator literature and the decision logic is internally consistent, this work could serve as a valuable practitioner guide that reduces misuse of IT measures in ML and agent-based AI. The provision of concrete artifacts like the flowchart and Bridge Boxes, along with domain coverage for both AI/ML and decision-making agents, strengthens its utility as a synthesis rather than a novel theoretical contribution.

major comments (2)
  1. [Abstract and §3] The selection of exactly these seven measures as the core set is presented without explicit justification for why other IT measures (e.g., conditional mutual information or directed information) are excluded; this choice is load-bearing for the framework's completeness claim.
  2. [Bridge Boxes and worked examples] The Bridge Boxes are described as interpretive links, but in the worked examples, they appear to equate mathematical quantities directly to cognitive constructs without sufficient caveats on the strength of those mappings, which could mislead practitioners on inferential claims.
minor comments (2)
  1. [Master decision table] Ensure that all estimator recommendations in the master table include references to the original papers or standard implementations to allow readers to verify the listed failure modes.
  2. [Figure 1] The flowchart could benefit from clearer branching conditions for high-dimensional data to avoid ambiguity in practical use.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive review and positive recommendation for minor revision. We address each major comment below and will incorporate the suggested clarifications into the revised manuscript.

read point-by-point responses

  1. Referee: [Abstract and §3] The selection of exactly these seven measures as the core set is presented without explicit justification for why other IT measures (e.g., conditional mutual information or directed information) are excluded; this choice is load-bearing for the framework's completeness claim.

    Authors: We agree that an explicit justification for the selection of these seven measures would strengthen the presentation. The measures were chosen because they constitute the most commonly applied IT quantities across the two domains covered by the guide: standard ML tasks (entropy, cross-entropy, mutual information, transfer entropy) and agent-based complexity analysis (integrated information, effective information, autonomy). Variants such as conditional mutual information and directed information are important but function as extensions or special cases of the included base measures; including every possible variant would expand the framework beyond a practical decision guide. In the revision we will add a short dedicated paragraph in §3 that states the selection criteria, notes the scope, and explicitly acknowledges that the framework does not claim exhaustive coverage of all IT measures. revision: yes

  2. Referee: [Bridge Boxes and worked examples] The Bridge Boxes are described as interpretive links, but in the worked examples, they appear to equate mathematical quantities directly to cognitive constructs without sufficient caveats on the strength of those mappings, which could mislead practitioners on inferential claims.

    Authors: We appreciate this observation. The Bridge Boxes are intended as heuristic interpretive aids drawn from existing literature associations rather than as direct equivalences. To reduce the risk of over-interpretation, we will revise the introductory description of the Bridge Boxes and insert additional qualifying language in each of the three worked examples. Specifically, we will add explicit statements that the mappings are suggestive and context-dependent, include phrases such as “potentially linked to” or “interpretive association,” and remind readers of the inferential limitations. These changes will be made without altering the core content of the examples. revision: yes

Circularity Check

0 steps flagged

No significant circularity: synthesis of existing measures with no derivations or fitted predictions

full rationale

The paper is explicitly a practitioner decision guide and synthesis of seven established information-theoretic measures. It supplies a flowchart, master table, and interpretive Bridge Boxes without any new equations, first-principles derivations, parameter fitting, or predictive claims. No load-bearing step reduces to a self-definition, self-citation chain, or renamed known result; the central artifacts are organizational aids whose validity rests on external literature rather than internal construction. This is the expected outcome for a non-derivational survey paper.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The framework rests on the assumption that the seven measures are the primary ones relevant to AI and that standard interpretations plus the Bridge Boxes are sufficient for safe use.

axioms (2)
  • domain assumption The seven measures (entropy, cross-entropy, mutual information, transfer entropy, Phi, effective information, autonomy) cover the main IT quantities needed in AI and agent contexts.
    The abstract presents these as the focus without justifying exclusion of other measures.
  • domain assumption Bridge Boxes provide accurate links from IT quantities to cognitive constructs.
    Mentioned as part of the framework but not derived or validated in the abstract.

pith-pipeline@v0.9.0 · 5515 in / 1344 out tokens · 56462 ms · 2026-05-08T06:00:52.219783+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

56 extracted references · 32 canonical work pages · 2 internal anchors

  1. [1]

    Springer, p 5–15, https://doi.org/10.1007/ 978-3-319-75726-1_1

    Albantakis L (2018) A tale of two animats: What does it take to have goals? In: Wandering Towards a Goal. Springer, p 5–15, https://doi.org/10.1007/ 978-3-319-75726-1_1

    2018

  2. [2]

    Entropy 23(11):1415

    Albantakis L (2021) Quantifying the autonomy of structurally diverse automata: A comparison of candidate measures. Entropy 23(11):1415. https://doi.org/10. 3390/e23111415

    2021

  3. [3]

    Entropy 17(8):5472–5502

    AlbantakisL,TononiG(2015)Theintrinsiccause-effectpowerofdiscretedynam- ical systems: From elementary cellular automata to adapting animats. Entropy 17(8):5472–5502. https://doi.org/10.3390/e17085472

    work page doi:10.3390/e17085472 2015

  4. [4]

    Entropy 21(10):989

    Albantakis L, Tononi G (2019) Causal composition: Structural differences among dynamically equivalent systems. Entropy 21(10):989. https://doi.org/10.3390/ e21100989

    2019

  5. [5]

    PLOS Com- putational Biology18(9), 1010492 (2022) https://doi.org/10.1371/journal.pcbi

    Albantakis L, Olbrich E, Zanoci C, et al (2014) Evolution of integrated causal structures in animats exposed to environments of increasing complexity. PLoS Computational Biology 10(12):e1003966. https://doi.org/10.1371/journal.pcbi. 1003966

    work page doi:10.1371/journal.pcbi 2014

  6. [6]

    ALBANTAKIS, L

    Albantakis L, et al (2023) Integrated information theory (IIT) 4.0: Formulating the properties of phenomenal existence in physical terms. PLoS Computational Biology 19(10):e1011465. https://doi.org/10.1371/journal.pcbi.1011465

    work page doi:10.1371/journal.pcbi.1011465 2023

  7. [7]

    Entropy 26(5):387

    Álvarez Chaves M, Gupta HV, Ehret U, et al (2024) On the accurate estimation of information-theoretic quantities from multi-dimensional sample data. Entropy 26(5):387. https://doi.org/10.3390/e26050387

    work page doi:10.3390/e26050387 2024

  8. [8]

    Artificial Intelligence p 103954

    Ao Z, Li J (2023) Entropy estimation via uniformization. Artificial Intelligence p 103954. https://doi.org/10.1016/j.artint.2023.103954

    work page doi:10.1016/j.artint.2023.103954 2023

  9. [9]

    In: Proceedings of the 35th International Conference on Machine Learning (ICML)

    Belghazi MI, et al (2018) Mutual information neural estimation. In: Proceedings of the 35th International Conference on Machine Learning (ICML)

    2018

  10. [10]

    Neural Computation 13(11):2409–2463

    Bialek W, Nemenman I, Tishby N (2001) Predictability, complexity, and learning. Neural Computation 13(11):2409–2463. https://doi.org/10.1162/ 089976601753195969

    2001

  11. [11]

    Nature Neuroscience 2(11):947–957

    Borst A, Theunissen FE (1999) Information theory and neural coding. Nature Neuroscience 2(11):947–957. https://doi.org/10.1038/14731

    work page doi:10.1038/14731 1999

  12. [12]

    Friston, The free-energy principle: a unified brain theory?, Nature Reviews Neuroscience 11 (2010) 127–138

    Friston K (2010) The free-energy principle: A unified brain theory? Nature Reviews Neuroscience 11(2):127–138. https://doi.org/10.1038/nrn2787 21

    work page doi:10.1038/nrn2787 2010

  13. [13]

    In: Proceedings of the 33rd International Conference on Machine Learning (ICML)

    GalY, GhahramaniZ (2016)Dropout as a Bayesian approximation:Representing model uncertainty in deep learning. In: Proceedings of the 33rd International Conference on Machine Learning (ICML). PMLR, pp 1050–1059

    2016

  14. [14]

    In: Advances in Neural Information Processing Systems (NeurIPS)

    GhimireS,etal(2021)ReliableestimationofKLdivergenceusingadiscriminator in RKHS. In: Advances in Neural Information Processing Systems (NeurIPS)

    2021

  15. [15]

    Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor

    Haarnoja T, Zhou A, Abbeel P, et al (2018) Soft actor-critic: Off-policy maximum entropy deep reinforcement learning with a stochastic actor. In: Pro- ceedings of the 35th International Conference on Machine Learning (ICML), arXiv:1801.01290

    work page internal anchor Pith review arXiv 2018

  16. [16]

    In: NeurIPS Workshop on Deep Learning and Representation Learning

    Hinton G, Vinyals O, Dean J (2015) Distilling the knowledge in a neural network. In: NeurIPS Workshop on Deep Learning and Representation Learning

    2015

  17. [17]

    Learning deep representations by mutual information estimation and maximization

    Hjelm RD, Fedorov A, Lavoie-Marchildon S, et al (2019) Learning deep repre- sentations by mutual information estimation and maximization. In: International Conference on Learning Representations (ICLR), arXiv:1808.06670

    work page Pith review arXiv 2019

  18. [18]

    Quantifying causal emergence,

    Hoel EP, Albantakis L, Tononi G (2013) Quantifying causal emergence shows that macro can beat micro. Proceedings of the National Academy of Sciences 110(49):19790–19795. https://doi.org/10.1073/pnas.1314922110

    work page doi:10.1073/pnas.1314922110 2013

  19. [19]

    Journal of Open Source Software 3(25):738

    James RG, Ellison CJ, Crutchfield JP (2018)dit: A python package for discrete information theory. Journal of Open Source Software 3(25):738. https://doi.org/ 10.21105/joss.00738

    work page doi:10.21105/joss.00738 2018

  20. [20]

    ACM Transactions on Human-Robot Interaction https://doi.org/10.1145/3800955, just Accepted

    Jiang H, Croft EA, Burke MG (2024) Influence-based reward modulation for implicit communication in human-robot interaction. ACM Transactions on Human-Robot Interaction https://doi.org/10.1145/3800955, just Accepted

    work page doi:10.1145/3800955 2024

  21. [21]

    Auto-Encoding Variational Bayes

    Kingma DP, Welling M (2014) Auto-encoding variational Bayes. In: International Conference on Learning Representations (ICLR), arXiv:1312.6114

    work page internal anchor Pith review arXiv 2014

  22. [22]

    Estimating Mutual Information

    Kraskov A, Stögbauer H, Grassberger P (2004) Estimating mutual information. Physical Review E 69(6):066138. https://doi.org/10.1103/PhysRevE.69.066138

    work page doi:10.1103/physreve.69.066138 2004

  23. [23]

    The Annals of Mathematical Statistics 22(1):79–86

    Kullback S, Leibler RA (1951) On information and sufficiency. The Annals of Mathematical Statistics 22(1):79–86

    1951

  24. [24]

    In: Proceedings of the 37th International Conference on Machine Learning (ICML)

    Laskin M, Srinivas A, Abbeel P (2020) CURL: Contrastive unsupervised repre- sentations for reinforcement learning. In: Proceedings of the 37th International Conference on Machine Learning (ICML). PMLR, pp 5639–5650

    2020

  25. [25]

    arXiv preprint arXiv:210206234 22

    Lazic N, et al (2021) Optimization issues in KL-constrained approximate policy iteration. arXiv preprint arXiv:210206234 22

    2021

  26. [26]

    Frontiers in Robotics and AI 1:11

    Lizier JT (2014) JIDT: An information-theoretic toolkit for studying the dynam- ics of complex systems. Frontiers in Robotics and AI 1:11. https://doi.org/10. 3389/frobt.2014.00011

    work page arXiv 2014

  27. [27]

    Cognitive Neurodynamics 7(3):253–261

    Ma SA, et al (2013) Estimating causal interaction between prefrontal cortex and striatum by transfer entropy. Cognitive Neurodynamics 7(3):253–261. https: //doi.org/10.1007/s11571-012-9233-2

    work page doi:10.1007/s11571-012-9233-2 2013

  28. [28]

    Communications in Statistics – Simulation and Computation https://doi.org/10.1080/03610918.2024.2316371, arXiv:2406.19432

    Madukaife MS, Phuc HD (2024) Estimation of Shannon differential entropy: An extensive comparative review. Communications in Statistics – Simulation and Computation https://doi.org/10.1080/03610918.2024.2316371, arXiv:2406.19432

    work page doi:10.1080/03610918.2024.2316371 2024

  29. [29]

    PLoS Computational Biology 14(7):e1006343

    Mayner WGP, Marshall W, Albantakis L, et al (2018) PyPhi: A toolbox for integrated information theory. PLoS Computational Biology 14(7):e1006343. https://doi.org/10.1371/journal.pcbi.1006343

    work page doi:10.1371/journal.pcbi.1006343 2018

  30. [30]

    Measuring Integrated Information: Comparison of Candidate Measures in Theory and Simulation

    Mediano PAM, Seth AK, Barrett AB (2019) Measuring integrated information: Comparison of candidate measures in theory and simulation. Entropy 21(1):17. https://doi.org/10.3390/e21010017

    work page doi:10.3390/e21010017 2019

  31. [31]

    Frontiers in Robotics and AI 5:60

    Moore DG, Valentini G, Walker SI, et al (2018) Inform: Efficient information- theoretic analysis of collective behaviors. Frontiers in Robotics and AI 5:60. https: //doi.org/10.3389/frobt.2018.00060

    work page doi:10.3389/frobt.2018.00060 2018

  32. [32]

    Physical Review E 87(5):052108

    Pampu NC, et al (2013) Transfer entropy as a tool for reconstructing interaction delays in neural signals. Physical Review E 87(5):052108. https://doi.org/10. 1103/PhysRevE.87.052108

    2013

  33. [33]

    MIT Press

    Parr T, Pezzulo G, Friston KJ (2022) Active Inference: The Free Energy Principle in Mind, Brain, and Behavior. MIT Press

    2022

  34. [34]

    Cambridge University Press

    Pearl J (2009) Causality: Models, Reasoning, and Inference, 2nd edn. Cambridge University Press

    2009

  35. [35]

    arXiv preprint arXiv:190700325

    Perry R, Mehta R, Guo R, et al (2019) Random forests for adaptive near- est neighbor estimation of information-theoretic quantities. arXiv preprint arXiv:190700325

    2019

  36. [36]

    arXiv preprint arXiv:220601400

    Piras D, et al (2022) A robust estimator of mutual information for deep learning interpretability. arXiv preprint arXiv:220601400

    2022

  37. [37]

    Machine Learning 1(1):81–106

    Quinlan JR (1986) Induction of decision trees. Machine Learning 1(1):81–106. https://doi.org/10.1007/BF00116251

    work page doi:10.1007/bf00116251 1986

  38. [38]

    Rakelly K, Gupta A, Florensa C, et al (2021) Which mutual-information rep- resentation learning objectives are sufficient for control? In: Advances in Neural Information Processing Systems (NeurIPS), pp 26345–26357 23

    2021

  39. [39]

    Journal of Statistical Mechanics: Theory and Experiment 2019(12):124020

    Saxe AM, Bansal Y, Dapello J, et al (2019) On the information bottleneck the- ory of deep learning. Journal of Statistical Mechanics: Theory and Experiment 2019(12):124020. https://doi.org/10.1088/1742-5468/ab3985

    work page doi:10.1088/1742-5468/ab3985 2019

  40. [40]

    Schreiber, Measuring information transfer, Physical review letters 85 (2) (2000) 461

    Schreiber T (2000) Measuring information transfer. Physical Review Letters 85(2):461–464. https://doi.org/10.1103/PhysRevLett.85.461

    work page doi:10.1103/physrevlett.85.461 2000

  41. [41]

    Trust Region Policy Optimization

    Schulman J, Levine S, Abbeel P, et al (2015) Trust region policy optimization. In: Proceedings of the 32nd International Conference on Machine Learning (ICML), arXiv:1502.05477

    work page Pith review arXiv 2015

  42. [42]

    arXiv preprint arXiv:170706347

    Schulman J, Wolski F, Dhariwal P, et al (2017) Proximal policy optimization algorithms. arXiv preprint arXiv:170706347

    2017

  43. [43]

    Elham Tabassi

    Shannon CE (1948) A mathematical theory of communication. Bell System Tech- nical Journal 27(3):379–423. https://doi.org/10.1002/j.1538-7305.1948.tb01338. x

    work page doi:10.1002/j.1538-7305.1948.tb01338 1948

  44. [44]

    PLoS Computational Biology 17(4)

    Shorten DP, Spinney RE, Lizier JT (2021) Estimating transfer entropy in con- tinuous time between neural spike trains. PLoS Computational Biology 17(4). https://doi.org/10.1371/journal.pcbi.1008054

    work page doi:10.1371/journal.pcbi.1008054 2021

  45. [45]

    PLoS Compu- tational Biology 12(11):e1005123

    Tegmark M (2016) Improved measures of integrated information. PLoS Compu- tational Biology 12(11):e1005123. https://doi.org/10.1371/journal.pcbi.1005123

    work page doi:10.1371/journal.pcbi.1005123 2016

  46. [46]

    In: IEEE Information Theory Workshop (ITW), https://doi.org/10

    Tishby N, Zaslavsky N (2015) Deep learning and the information bottleneck principle. In: IEEE Information Theory Workshop (ITW), https://doi.org/10. 1109/ITW.2015.7133169

    work page arXiv 2015

  47. [47]

    In: Proceedings of the 37th Annual Allerton Conference on Communication, Control, and Computing

    Tishby N, Pereira FC, Bialek W (2000) The information bottleneck method. In: Proceedings of the 37th Annual Allerton Conference on Communication, Control, and Computing

    2000

  48. [48]

    Nature Reviews Neuroscience 17(7):450–

    TononiG,BolyM,MassiminiM,etal(2016)Integratedinformationtheory:From consciousness to its physical substrate. Nature Reviews Neuroscience 17(7):450–

    2016

  49. [49]

    https://doi.org/10.1038/nrn.2016.44

    work page doi:10.1038/nrn.2016.44 2016

  50. [50]

    In: International Conference on Learning Representations (ICLR)

    Tschannen M, Djolonga J, Rubenstein PK, et al (2020) On mutual informa- tion maximization for representation learning. In: International Conference on Learning Representations (ICLR)

    2020

  51. [51]

    Machine Learning: Science and Technology https://doi.org/10.1088/2632-2153/ac9455

    Vaitl L, et al (2022) Gradients should stay on path: Better estimators of the reverse- and forward KL divergence for normalizing flows. Machine Learning: Science and Technology https://doi.org/10.1088/2632-2153/ac9455

    work page doi:10.1088/2632-2153/ac9455 2022

  52. [52]

    arXiv preprint arXiv:180703748 24

    van den Oord A, Li Y, Vinyals O (2018) Representation learning with contrastive predictive coding. arXiv preprint arXiv:180703748 24

    2018

  53. [53]

    https: //github.com/gregversteeg/NPEET, python library for entropy and mutual information estimation

    Ver Steeg G (2014) NPEET: Non-parametric entropy estimation toolbox. https: //github.com/gregversteeg/NPEET, python library for entropy and mutual information estimation

    2014

  54. [54]

    Journal of Computational Neuroscience 30(1):45–67

    Vicente R, Wibral M, Lindner M, et al (2011) Transfer entropy – a model-free measure of effective connectivity for the neurosciences. Journal of Computational Neuroscience 30(1):45–67. https://doi.org/10.1007/s10827-010-0262-3

    work page doi:10.1007/s10827-010-0262-3 2011

  55. [55]

    Springer, https://doi.org/10.1007/978-3-642-54474-3

    Wibral M, Vicente R, Lizier JT (eds) (2014) Directed Information Measures in Neuroscience. Springer, https://doi.org/10.1007/978-3-642-54474-3

    work page doi:10.1007/978-3-642-54474-3 2014

  56. [56]

    Journal of Open Source Software 4(34):1081

    Wollstadt P, et al (2019) IDTxl: The information dynamics toolkit xl: A python package for the efficient analysis of multivariate information dynamics in net- works. Journal of Open Source Software 4(34):1081. https://doi.org/10.21105/ joss.01081 25

    2019