Pith. sign in

REVIEW 3 major objections 4 minor 299 references

This paper proposes a formal way to describe any finite dynamical system as a chemistry of tokens and reactions, and argues that Game of Life gliders satisfy the added uniqueness criterion.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-01 13:27 UTC pith:OE74T2DU

load-bearing objection Genuinely new formal framework for describing dynamical systems as chemistries, but the main glider example has a concrete arithmetic bug in eq. (35) that breaks the claimed compatibility; the definitions are worth a serious look, the demonstration needs fixing. the 3 major comments →

arxiv 2607.19090 v1 pith:OE74T2DU submitted 2026-07-21 cs.DM cs.NEnlin.PS

Towards chemistries in dynamical systems

Martin Biehl , Nathaniel Virgo This is my paper
classification cs.DM cs.NEnlin.PS MSC 68Q8037B15
keywords pattern chemistrydynamical systemsPetri netschemical reaction networksGame of Lifegliderscellular automataunambiguous transitions
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The authors define a 'pattern chemistry' as a way to pick out tokens (molecules) in a dynamical system, choose a Petri-net of reactions, and assign a transition map that turns the tokens in one state into the tokens in the next. They require this assignment to be compatible with the system's update rule, and add a criterion: the transition map should always choose the unique smallest set of reactions. They show that a glider in the finite Game of Life passes this criterion, so its movement can be described as a single 'movement' reaction rather than annihilation plus creation. The paper treats this as an exploratory step toward a formal account of objects like particles, organisms, or patterns in complex systems.

Core claim

The central discovery is a working definition: a pattern chemistry of a dynamical system (X,h) consists of (i) an indexed family of functions that read off tokens from each state, (ii) a Petri-net of reactions on token types, and (iii) a transition map θ that for every state x returns a step whose source multiset equals the tokens in x and whose target multiset equals the tokens in h(x). The paper then strengthens this to an 'unambiguous' pattern chemistry by demanding that θ(x) be the unique step with the fewest transitions. The main positive example is the glider in the finite Game of Life: with movement transitions defined via light cones, the minimum-step solution is always unique, so th

What carries the argument

The central objects are (a) the token map ϕ, which turns a state into a multiset of tokens by counting pattern shapes in each frame of reference; (b) the Petri-net whose transitions have source and target multisets; and (c) the transition map θ, which picks the step realizing the update. The decisive mechanism is the 'light cone' construction: for each glider token at position p, the frame of reference at the next time is any p' in LC(p), the set of positions a glider could legitimately move to. The uniqueness criterion (minimizing transition count) gives a principled way to choose movement transitions over annihilation+creation, and the paper argues the glider's light-cone structure guarant

Load-bearing premise

The glider example assumes that at most one glider ever appears in a single light cone at the next time step, so that the products in the transition map act as logical yes/no gates.

What would settle it

Find a finite Game of Life configuration in which two distinct glider tokens in the next state lie in the light cone of a single glider in the current state (or two gliders in the current state share one possible successor in a light cone). If such a state exists on the finite torus, the minimum-transition step is ambiguous and the glider chemistry fails the unambiguous criterion.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • If correct, the definition gives a formal bridge between dynamical systems and chemical reaction networks, allowing chemistry-based reasoning to be applied to any finite-state dynamics.
  • The unambiguous criterion provides a formal way to say when a pattern is 'really' a moving object: exactly when the fewest-reactions description is unique.
  • The paper's examples show that naive choices (cells as tokens) are ambiguous, while gliders are not, supporting the intuition that gliders are natural individuals.
  • The framework applies to any finite dynamical system, not just cellular automata, so it can be used to ask what additional structure a cellular automaton provides.
  • The authors identify open problems: avoiding degenerate choices of frame-of-reference functions and finding practical algorithms for discovering interesting chemistries.

Where Pith is reading between the lines

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

  • One could test the uniqueness criterion on other Game of Life objects (eaters, blinkers, other spaceships) to see whether the criterion picks out exactly the patterns that human observers treat as mobile individuals.
  • The light-cone uniqueness argument might be recast as a no-cloning/no-merging condition on the pattern's local dynamics; if patterns can branch or fuse, the criterion fails, suggesting a formal notion of 'object identity' in discrete systems.
  • The definition suggests a computationally tractable route: start with annihilation/creation transitions and greedily add movement (or other) transitions that preserve uniqueness; this could be automated as a search over Petri-nets.
  • For infinite cellular automata, finite-support multisets fail; a limiting or profinite construction on increasing finite grids might recover glider chemistries on the infinite plane.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper proposes a formal framework for describing aspects of a finite dynamical system in chemical terms. It defines a 'pattern chemistry' as a token map from states to multisets of (frame, shape) pairs, a Petri-net whose places are these pairs, and a transition map assigning a step (multiset of transitions) to each state; compatibility requires that the step's source multiset equals the current token multiset and its target multiset equals the token multiset after one update. The paper adds an 'unambiguous' criterion: among all steps realizing a given token transition, the step with the fewest transitions must be unique. The main worked example is a glider in finite Game of Life, for which the authors attempt to construct a chemistry with movement, annihilation, and creation transitions and argue that it satisfies the unambiguity criterion. The paper also discusses alternative choices and limitations.

Significance. The definitions are self-contained and the proposed framework is a genuine attempt to formalize pattern movement in cellular automata as chemical reactions. If the specific construction in the glider example can be repaired, the paper would provide a useful conceptual tool for artificial life and for relating dynamical-system descriptions to Petri-net/chemical-reaction-network descriptions. The unambiguity criterion is a novel, non-circular design principle and is clearly distinguished from simple curve fitting. The paper is exploratory rather than a complete theory, but the formal backbone is coherent. No machine-checked proofs or reproducible code are provided; the value lies in the definitions and the illustrative example.

major comments (3)
  1. [Definition 8 (unambiguous pattern chemistry)] The minimality condition is written as |ϕ(x)| ≤ |t′| and equality implies t′ = ϕ(x). But |ϕ(x)| is the number of tokens in the initial marking, not the number of transitions in a step. The intended criterion, stated in prose, is that θ(x) should be a step with the fewest transitions; the comparison should be |θ(x)| ≤ |t′| for all t′ with src(t′)=ϕ(x), tgt(t′)=ϕ(h(x)). As written, Definition 8 does not express the stated property. This is a load-bearing error in the central definition.
  2. [Eq. (35), glider chemistry construction] The formula for θ̄(x) does not satisfy the compatibility condition (18)–(19). For a fixed p with ϕ(x)(p)=1, the second sum contributes one annihilation transition |†,p⟩ for every p′∈LC(p) with ϕ(h(x))(p′)=0, and the first sum contributes a movement transition for every occupied p′. Thus if LC(p) contains k positions, the total source contribution from p is k|p⟩ (k−m movement/annihilation copies plus m movement copies), not |p⟩. In particular, when no p′ in LC(p) is occupied, the step includes k annihilations, so src(θ̄(x)) = k|p⟩ ≠ ϕ(x). The prose about unambiguous movement choices addresses a different issue and does not repair the arithmetic. The glider example is therefore not established as a pattern chemistry as written.
  3. [Section 'The unambiguous glider chemistry with movement'] Even if Eq. (35) is repaired, the claim that no ambiguity arises between possible movement transitions is asserted with 'it is not hard to see by trying it out' rather than proved. For the example to support the unambiguity criterion, the authors need a rigorous argument, or a finite exhaustive check for the chosen grid size, showing that in every state x there is at most one occupied p′∈LC(p) for each p, and no alternative step with the same number of transitions exists. This is load-bearing for the paper's central demonstration.
minor comments (4)
  1. [Abstract] The sentence 'We discuss gliders in the game of life cellular and argue that when following their definition of according to Randall Beer' appears to have missing words; it should likely read 'gliders in the Game of Life cellular automaton' and 'following the definition of gliders according to Randall Beer'.
  2. [After Eq. (35)] The parenthetical 'note the use of (1− ¯ϕ(x)) for later' seems to contain a typo; the expression should refer to (1−ϕ(h(x))(p′)) or an analogous indicator, not ¯ϕ(x).
  3. [Commutative diagram in Section 'Pattern chemistries of a dynamical systems'] The diagram is not typeset in a way that makes the direction of the src and tgt arrows clear; consider using a labeled commutative diagram environment.
  4. [Example 5] The phrase 'we then find that θ:X→M(n×n) just returns the multiset' is confusing because the transition map should return a multiset of transitions, not a multiset of places; likely a typo for ϕ.

Circularity Check

0 steps flagged

No significant circularity: the framework is defined constructively and the glider example is an explicit worked illustration, not a fitted prediction or a self-citation load-bearing result.

full rationale

The paper's central contribution is a set of definitions (Definitions 7 and 8) plus an explicit construction for gliders in a finite Game of Life. The compatibility conditions (18)-(19) are requirements the transition map must satisfy, not quantities fitted from data. The transition map in eq. (35) is written out directly in terms of the token map and light cones; it is not obtained by solving for a parameter that is then renamed as a prediction. The unambiguity argument appeals to geometric facts about glider light cones and the minimization criterion is proposed as a desideratum, not imported from a self-citation. Citations to Beer (2014, 2020a) supply glider conventions and the light-cone idea, but they are external and not used to justify the formal framework itself. The paper even flags open issues and limitations. A possible internal inconsistency in eq. (35) would be a correctness or proof gap, not circular reasoning, because it does not make the claimed result equivalent to its assumptions by construction. No fitted-input-called-prediction, self-citation chain, or renaming pattern is present.

Axiom & Free-Parameter Ledger

2 free parameters · 4 axioms · 2 invented entities

The paper introduces no fitted physical constants or data. Its load-bearing choices are the observer's hand-picked frame/shape family and light-cone relation, plus the domain assumption of finite toroidal GoL. The most fragile item is the unproved uniqueness/light-cone non-overlap assertion and the arithmetic behavior of eq. (35).

free parameters (2)
  • R-indexed family {fr} (frames and shapes) = R = symmetry group G of finite GoL; S = {rocket, wedge}
    Hand-chosen in the glider example; the paper notes that other choices are possible and that degeneracy is an open problem.
  • Light cone relation LC(g,s) = imported from Beer (2014)
    Defines movement transitions; the paper assumes the relation makes eq. (35)'s sums behave as yes/no indicators.
axioms (4)
  • domain assumption Multisets have finite support and only finite dynamical systems are considered.
    Stated in the introduction and Definition 1; the entire construction (token maps, steps, uniqueness) relies on finite multiset arithmetic.
  • domain assumption The Game of Life is taken on a finite n×n grid with periodic boundary conditions.
    Given in Example 1; the glider light-cone constructions and finite-state arguments depend on this domain.
  • domain assumption The unambiguity criterion (unique least number of transitions) is adopted as the desired property of a pattern chemistry.
    Proposed in Definition 8 and motivated by observer parsimony; not derived from the dynamics. If one rejects this criterion, the main example's significance changes.
  • ad hoc to paper For every state x, no two glider tokens' light cones cause multiple minimal matchings; in particular eq. (35)'s sum-of-complements behaves as an existence test.
    This is asserted informally in the glider example and is not proved or exhaustively checked; the unambiguous glider chemistry rests on it.
invented entities (2)
  • pattern tokens (elements of M(R×S)) no independent evidence
    purpose: Represent observable patterns as chemical species in the Petri-net.
    Formal observables defined by the observer; no independent empirical signature.
  • movement transitions p→p′ no independent evidence
    purpose: Allow one token to be transformed into another within the light cone instead of annihilation-plus-creation; central to unambiguity.
    Added by hand in the glider chemistry; not derived from the dynamics.

pith-pipeline@v1.3.0-alltime-deepseek · 11501 in / 15607 out tokens · 144879 ms · 2026-08-01T13:27:52.721133+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Towards chemistries in dynamical systems." pith.science (2026). https://pith.science/paper/OE74T2DU

@misc{pith2026260719090,
  author       = {Pith},
  title        = {Pith review of: Towards chemistries in dynamical systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OE74T2DU}},
  note         = {Machine review of arXiv:2607.19090}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Chemistry describes aspects of the universe in terms of molecules and their reactions. In this exploratory work we present a way to describe aspects of any dynamical system in similar terms. To describe a dynamical system in this way three decisions have to be made. The first is how many different "places" there are at which molecules or chemical species can occur; the second is how to determine the species present (or not) at each place; and the third is the set of transitions and reactions that can occur between the species in the various places. For these choices to be compatible with the state update of the dynamical system each state must be able to determine transitions that take the currently occurring molecules to those occurring in the updated state. We also propose an additional requirement that there is always a unique way to choose the least amount of transitions occurring during state updates. We discuss gliders in the game of life cellular and argue that when following their definition of according to Randall Beer they satisfy the additional criterion as well. We also point out some issues with the approach.

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

299 extracted references · 67 canonical work pages · 2 internal anchors

  1. [1]

    Adaptive Behavior , volume=

    The necessity of extended autopoiesis , author=. Adaptive Behavior , volume=. 2020 , publisher=

    2020

  2. [2]

    Advances in Artificial Life: 9th European Conference, ECAL 2007, Lisbon, Portugal, September 10-14, 2007

    Autonomy: a review and a reappraisal , author=. Advances in Artificial Life: 9th European Conference, ECAL 2007, Lisbon, Portugal, September 10-14, 2007. Proceedings 9 , pages=. 2007 , organization=

    2007

  3. [3]

    European Conference on Artificial Life , pages=

    The role of the spatial boundary in autopoiesis , author=. European Conference on Artificial Life , pages=. 2009 , organization=

    2009

  4. [4]

    Artificial Life Conference Proceedings , pages=

    An active inference implementation of phototaxis , author=. Artificial Life Conference Proceedings , pages=. 2017 , organization=

    2017

  5. [5]

    2020 International Joint Conference on Neural Networks (IJCNN) , pages=

    A Bayesian perspective on classical control , author=. 2020 International Joint Conference on Neural Networks (IJCNN) , pages=. 2020 , organization=

    2020

  6. [6]

    Artificial life conference proceedings , pages=

    The modularity of action and perception revisited using control theory and active inference , author=. Artificial life conference proceedings , pages=. 2018 , organization=

    2018

  7. [7]

    Artificial Life Conference Proceedings 32 , pages=

    Predictions in the eye of the beholder: An active inference account of Watt governors , author=. Artificial Life Conference Proceedings 32 , pages=. 2020 , organization=

    2020

  8. [8]

    arXiv preprint arXiv:1503.04187 , year=

    A minimal active inference agent , author=. arXiv preprint arXiv:1503.04187 , year=

    Pith/arXiv arXiv

  9. [9]

    Lavore, Elena Di and Román, Mario and Sobociński, Paweł , month = jan, year =. Partial. doi:10.48550/arXiv.2502.03477 , abstract =

    doi:10.48550/arxiv.2502.03477

  10. [10]

    Jacobs, Bart and Kissinger, Aleks and Zanasi, Fabio , month = jul, year =. Causal. doi:10.48550/arXiv.1811.08338 , abstract =

    Pith/arXiv arXiv doi:10.48550/arxiv.1811.08338

  11. [11]

    Retrospective on

    Miyazono, Evan and Lakin, Chris , month = may, year =. Retrospective on

  12. [12]

    Formalising the intentional stance 1: attributing goals and beliefs to stochastic processes

    McGregor, Simon and timorl and Virgo, Nathaniel , month = jan, year =. Formalising the intentional stance 1: attributing goals and beliefs to stochastic processes , shorttitle =. doi:10.48550/arXiv.2405.16490 , abstract =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2405.16490

  13. [13]

    Formalising the intentional stance 2: a coinductive approach

    McGregor, Simon and timorl and Virgo, Nathaniel , month = jan, year =. Formalising the intentional stance 2: a coinductive approach , shorttitle =. doi:10.48550/arXiv.2501.09173 , abstract =

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.2501.09173

  14. [14]

    Nathaniel Virgo , month = mar, year =. Sussex

  15. [15]

    good regulator theorem

    Nathaniel Virgo , month = mar, year =. Extending the “good regulator theorem” -

  16. [16]

    Embedded

    Demski, Abram and Garrabrant, Scott , month = oct, year =. Embedded. doi:10.48550/arXiv.1902.09469 , abstract =

    Pith/arXiv arXiv doi:10.48550/arxiv.1902.09469 1902

  17. [17]

    2023 , eprint=

    A Causal Analysis of Harm , author=. 2023 , eprint=

    2023

  18. [18]

    2022 , eprint=

    Quantifying Harm , author=. 2022 , eprint=

    2022

  19. [19]

    Journal of Machine Learning Research , author =

    The d-. Journal of Machine Learning Research , author =. 2023 , pages =

    2023

  20. [20]

    Particle

    Mordvintsev, Alexander and Niklasson, Eyvind and Randazzo, Ettore , file =. Particle. 2022 , month = dec, url =

    2022

  21. [21]

    2022 , MONTH = Jan, HAL_ID =

    Learning Sensorimotor Agency in Cellular Automata , AUTHOR =. 2022 , MONTH = Jan, HAL_ID =

    2022

  22. [22]

    Andrew Critch , month = oct, year =

  23. [23]

    and Saint-Pierre, Patrick , year =

    Aubin, Jean-Pierre and Bayen, Alexandre M. and Saint-Pierre, Patrick , year =. Viability

  24. [24]

    2023 , journal =

    Categorical systems theory , author=. 2023 , journal =

    2023

  25. [25]

    GitHub repository , howpublished =

    Capucci, Matteo , title =. GitHub repository , howpublished =. 2025 , publisher =

    2025

  26. [26]

    arXiv preprint arXiv:2405.06624 , year=

    Towards Guaranteed Safe AI: A Framework for Ensuring Robust and Reliable AI Systems , author=. arXiv preprint arXiv:2405.06624 , year=

    Pith/arXiv arXiv

  27. [27]

    International Conference on Machine Learning , pages=

    A Category-theoretical Meta-analysis of Definitions of Disentanglement , author=. International Conference on Machine Learning , pages=. 2023 , organization=

    2023

  28. [28]

    International journal of systems science , volume=

    Every good regulator of a system must be a model of that system , author=. International journal of systems science , volume=. 1970 , publisher=

    1970

  29. [29]

    IEEE Transactions on Automatic Control , volume=

    Error feedback and internal models on differentiable manifolds , author=. IEEE Transactions on Automatic Control , volume=. 1984 , publisher=

    1984

  30. [30]

    Advances in Complex Systems , author =

    A. Advances in Complex Systems , author =. 2009 , pages =. doi:10.1142/S0219525909002155 , number =

    work page doi:10.1142/s0219525909002155 2009

  31. [31]

    Artificial Life , author =

    The cognitive domain of a glider in the game of life , volume =. Artificial Life , author =. 2014 , pmid =. doi:10.1162/ARTL\_a\_00125 , abstract =

    doi:10.1162/artl 2014

  32. [32]

    Prediction and modularity in dynamical systems

    A. Kolchinsky and L. M. Rocha , title = "Prediction and modularity in dynamical systems", journal = "Advances in Artificial Life, ECAL", year = "2011", month = "08/08/2011", pages = "423--430", publisher = "MIT Press", address = "Paris, France", isbn = "0-262-29714-0", attachments = "http://dev.iu-pti.org/sites/default/files/0262297140chap65.pdf", editor ...

    arXiv 2011

  33. [33]

    Integrated Information in Discrete Dynamical Systems: Motivation and Theoretical Framework , volume =

    Balduzzi, David and Tononi, Giulio , month = jun, year =. Integrated Information in Discrete Dynamical Systems: Motivation and Theoretical Framework , volume =. doi:10.1371/journal.pcbi.1000091 , abstract =

    work page doi:10.1371/journal.pcbi.1000091

  34. [34]

    Advances in Complex Systems , author =

    A. Advances in Complex Systems , author =. 2008 , pages =. doi:10.1142/S0219525908001507 , number =

    work page doi:10.1142/s0219525908001507 2008

  35. [35]

    Neural Netw

    Dynamical properties of strongly interacting Markov chains , volume =. Neural Netw. , author =. 2003 , keywords =. doi:10.1016/S0893-6080(03)00190-4 , abstract =

    work page doi:10.1016/s0893-6080(03)00190-4 2003

  36. [36]

    and Snell, J

    Kemeny, John G. and Snell, J. Laurie , month = jul, year =. Finite Markov Chains: With a New Appendix

  37. [37]

    Biosystems , author =

    Autonomy: An information theoretic perspective , volume =. Biosystems , author =. 2008 , keywords =. doi:10.1016/j.biosystems.2007.05.018 , abstract =

    work page doi:10.1016/j.biosystems.2007.05.018 2008

  38. [38]

    Measuring autonomy and emergence via granger causality , volume =. Artif. Life , author =. 2010 , keywords =. doi:10.1162/artl.2010.16.2.16204 , abstract =

    work page doi:10.1162/artl.2010.16.2.16204 2010

  39. [39]

    Defining Emergent Descriptions by Information Preservation , journal =

    Polani, Daniel , year =. Defining Emergent Descriptions by Information Preservation , journal =

  40. [40]

    Emergence, Intrinsic Structure of Information, and Agenthood , copyright =

    Polani, Daniel , editor =. Emergence, Intrinsic Structure of Information, and Agenthood , copyright =. Unifying Themes in Complex Systems , publisher =. 2008 , keywords =

    2008

  41. [41]

    Blind construction of optimal nonlinear recursive predictors for discrete sequences , isbn =

    Shalizi, Cosma Rohilla and Shalizi, Kristina Lisa , year =. Blind construction of optimal nonlinear recursive predictors for discrete sequences , isbn =. Proceedings of the 20th conference on Uncertainty in artificial intelligence , publisher =

  42. [42]

    and Ellison, Christopher J

    Still, Susanne and Crutchfield, James P. and Ellison, Christopher J. , month = aug, year =. Optimal Causal Inference: Estimating Stored Information and Approximating Causal Architecture , shorttitle =

  43. [43]

    The 37th annual Allerton Conference on Communication, Control, and Computing , author =

    The Information Bottleneck Method , abstract =. The 37th annual Allerton Conference on Communication, Control, and Computing , author =. 1999 , pages =

    1999

  44. [44]

    Artificial Intelligence , author =

    A dynamical systems perspective on agent-environment interaction , volume =. Artificial Intelligence , author =. 1995 , pages =. doi:10.1016/0004-3702(94)00005-L , number =

    work page doi:10.1016/0004-3702(94)00005-l 1995

  45. [45]

    Theory in Biosciences , author =

    Information-driven self-organization: the dynamical system approach to autonomous robot behavior , volume =. Theory in Biosciences , author =. 2012 , keywords =. doi:10.1007/s12064-011-0137-9 , abstract =

    work page doi:10.1007/s12064-011-0137-9 2012

  46. [46]

    Organization of the information flow in the perception-action loop of evolved agents , doi =

    Klyubin,. Organization of the information flow in the perception-action loop of evolved agents , doi =. 2004. 2004 , keywords =

    2004

  47. [47]

    Information geometry on complexity and stochastic interaction , url =

    Ay, Nihat , month = dec, year =. Information geometry on complexity and stochastic interaction , url =. MPI MiS Preprint , note =

  48. [48]

    and Varela, Francisco J

    Maturana, Humberto R. and Varela, Francisco J. , year =. Autopoiesis and cognition: the realization of the living , isbn =

  49. [49]

    Empowerment: a universal agent-centric measure of control , volume =

    Klyubin,. Empowerment: a universal agent-centric measure of control , volume =. The 2005. 2005 , keywords =. doi:10.1109/CEC.2005.1554676 , abstract =

    arXiv 2005

  50. [50]

    Physical Review Letters , author =

    Causal Entropic Forces , volume =. Physical Review Letters , author =. 2013 , pages =. doi:10.1103/PhysRevLett.110.168702 , abstract =

    doi:10.1103/physrevlett.110.168702 2013

  51. [51]

    , month = apr, year =

    Oudeyer, Pierre-Yves and Kaplan, Frederic and Hafner, Verena V. , month = apr, year =. Intrinsic Motivation Systems for Autonomous Mental Development , volume =. doi:10.1109/TEVC.2006.890271 , number =

    arXiv 2006

  52. [52]

    A possibility for implementing curiosity and boredom in model-building neural controllers , isbn =

    Schmidhuber, Jürgen , year =. A possibility for implementing curiosity and boredom in model-building neural controllers , isbn =. Proceedings of the first international conference on simulation of adaptive behavior on From animals to animats , publisher =

  53. [53]

    Theoretical biology , language =

    von Uexk\"ull, Jakob , year =. Theoretical biology , language =

  54. [54]

    Artificial Life , booktitle =

    Langton, Christopher , editor =. Artificial Life , booktitle =. 1989 , pages =

    1989

  55. [55]

    Some ways to see two in one , isbn =

    Biehl, Martin and Polani, Daniel , month = sep, year =. Some ways to see two in one , isbn =. Advances in Artificial Life,. doi:10.7551/978-0-262-31709-2-ch165 , urldate =

    work page doi:10.7551/978-0-262-31709-2-ch165

  56. [56]

    Journal of Artificial Societies and Social Simulation , ISSN =

    Digested Information as an Information Theoretic Motivation for Social Interaction , author =. Journal of Artificial Societies and Social Simulation , ISSN =. 2011 , URL =

    2011

  57. [57]

    MPI MIS PREPRINT 95/2001 , year =

    Nihat Ay , title =. MPI MIS PREPRINT 95/2001 , year =

    2001

  58. [58]

    Kybernetes , author =

    Inside embodiment -- what means embodiment to radical constructivists? , volume =. Kybernetes , author =. 2005 , keywords =. doi:10.1108/03684920510575762 , number =

    work page doi:10.1108/03684920510575762 2005

  59. [59]

    Adaptive Behavior , author =

    Higher coordination with less control—a result of information maximization in the sensorimotor loop , volume =. Adaptive Behavior , author =. 2010 , pages =

    2010

  60. [60]

    Causality: models, reasoning, and inference , isbn =

    Pearl, Judea , year =. Causality: models, reasoning, and inference , isbn =

  61. [61]

    Advances in Complex Systems , author =

    Comparison between different methods of level identification , issn =. Advances in Complex Systems , author =. 2014 , pages =. doi:10.1142/S0219525914500076 , abstract =

    work page doi:10.1142/s0219525914500076 2014

  62. [62]

    Information and closure in systems theory , booktitle =

    Bertschinger, Nils and Olbrich, Eckehard and Ay, Nihat and Jost, J\"urgen , year =. Information and closure in systems theory , booktitle =

  63. [63]

    Advances in Complex Systems , author =

    Information Flows in Causal Networks , volume =. Advances in Complex Systems , author =. 2008 , pages =. doi:10.1142/S0219525908001465 , language =

    work page doi:10.1142/s0219525908001465 2008

  64. [64]

    Physical Review Letters , author =

    Measuring Information Transfer , volume =. Physical Review Letters , author =. 2000 , pages =. doi:10.1103/PhysRevLett.85.461 , abstract =

    doi:10.1103/physrevlett.85.461 2000

  65. [65]

    Physica A: Statistical Mechanics and its Applications , author =

    Information-theoretic approach to the study of control systems , volume =. Physica A: Statistical Mechanics and its Applications , author =. 2004 , keywords =. doi:10.1016/j.physa.2003.09.007 , abstract =

    work page doi:10.1016/j.physa.2003.09.007 2004

  66. [66]

    M and Thomas, Joy A , year =

    Cover, T. M and Thomas, Joy A , year =. Elements of information theory , isbn =

  67. [67]

    Neuroinformatics , author =

    Methods for Quantifying the Informational Structure of Sensory and Motor Data , volume =. Neuroinformatics , author =. 2005 , pages =. doi:10.1385/NI:3:3:243 , number =

    work page doi:10.1385/ni:3:3:243 2005

  68. [68]

    and Beer, Randall D

    Williams, Paul L. and Beer, Randall D. , year =. Information dynamics of evolved agents , isbn =. Proceedings of the 11th international conference on Simulation of adaptive behavior: from animals to animats , publisher =

  69. [69]

    Entropy , author =

    Quantifying Morphological Computation , volume =. Entropy , author =. 2013 , note =. doi:10.3390/e15051887 , abstract =

    work page doi:10.3390/e15051887 2013

  70. [70]

    , month = mar, year =

    Mitchell, Melanie and Hraber, Peter and Crutchfield, James P. , month = mar, year =. Revisiting the Edge of Chaos: Evolving Cellular Automata to Perform Computations , shorttitle =

  71. [71]

    Physica D: Nonlinear Phenomena , author =

    Computation at the edge of chaos: Phase transitions and emergent computation , volume =. Physica D: Nonlinear Phenomena , author =. 1990 , pages =. doi:10.1016/0167-2789(90)90064-V , abstract =

    doi:10.1016/0167-2789(90)90064-v 1990

  72. [72]

    and Prokopenko, Mikhail and Zomaya, Albert Y

    Lizier, Joseph T. and Prokopenko, Mikhail and Zomaya, Albert Y. , editor =. A Framework for the Local Information Dynamics of Distributed Computation in Complex Systems , copyright =. Guided Self-Organization: Inception , publisher =. 2014 , keywords =

    2014

  73. [73]

    Physical Review E , author =

    Fluctuation theorem and large deviation function for a solvable model of a molecular motor , volume =. Physical Review E , author =. 2008 , pages =. doi:10.1103/PhysRevE.78.011915 , abstract =

    work page doi:10.1103/physreve.78.011915 2008

  74. [74]

    Generalised Thermostatistics , isbn =

    Naudts, Jan , month = feb, year =. Generalised Thermostatistics , isbn =

  75. [75]

    Scandinavian Actuarial Journal , author =

    An overview of the theory of large deviations and applications to statistical mechanics , volume =. Scandinavian Actuarial Journal , author =. 1995 , pages =. doi:10.1080/03461238.1995.10413952 , abstract =

    arXiv 1995

  76. [76]

    Communications on Pure and Applied Mathematics , author =

    Asymptotic evaluation of certain markov process expectations for large time. Communications on Pure and Applied Mathematics , author =. 1983 , pages =. doi:10.1002/cpa.3160360204 , language =

    doi:10.1002/cpa.3160360204 1983

  77. [77]

    The Annals of Probability , author =

    Large Deviations and Maximum Entropy Principle for Interacting Random Fields on. The Annals of Probability , author =. 1993 , OPTnote =. doi:10.1214/aop/1176989002 , abstract =

    arXiv 1993

  78. [78]

    The Annals of Probability , author =

    Large Deviations for the Empirical Field of a Gibbs Measure , volume =. The Annals of Probability , author =. 1988 , note =. doi:10.1214/aop/1176991671 , abstract =

    arXiv 1988

  79. [79]

    Physics Reports , author =

    The large deviation approach to statistical mechanics , volume =. Physics Reports , author =. 2009 , note =. doi:10.1016/j.physrep.2009.05.002 , abstract =

    doi:10.1016/j.physrep.2009.05.002 2009

  80. [80]

    Mathematische Annalen , author =

    Untersuchungen \"uber die. Mathematische Annalen , author =. 1909 , keywords =. doi:10.1007/BF01450409 , language =

    doi:10.1007/bf01450409 1909

Showing first 80 references.