Pith. sign in

REVIEW 5 major objections 6 minor 81 references

Consciousness as a Functor

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read This paper proposes that consciousness is a functor carrying contents from an unconscious topos of coalgebras into conscious short-term memory, with MUMBLE as the topos's internal language of thought.

desk verdict A programmatic position paper that packages standard topos theory and coalgebra as a theory of consciousness, but the central topos-of-unconscious-processes claim is posited, not proven, and the load-bearing theorem is imported from a self-cited preprint. read the letter →

arxiv 2508.17561 v1 pith:7CGGBM3M submitted 2025-08-25 cs.AI cs.LG

Sridhar Mahadevan This is my paper · ORCID
classification cs.AIcs.LG
keywords consciousnessfunctortoposcoalgebraMUMBLEuniversalreinforcementlearningnetworkeconomicsasynchronousdistributedcomputation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper proposes that consciousness is best understood as a structured mapping—a functor—that carries contents from a large, parallel, unconscious memory system into a small, sequential, conscious short-term memory. Its central move is to model the ensemble of unconscious processes as a topos of coalgebras, which is a 'set-like' category admitting limits, exponentials, and a subobject classifier; every such topos carries an internal logic. The paper identifies that internal logic, called MUMBLE (Multi-modal Universal Mitchell-Bénabou Language Embedding), with the 'language of thought,' and gives it a Kripke–Joyal semantics. Information flow in the two directions is then handled by two mechanisms: Universal Reinforcement Learning compiles conscious trial-and-error into long-term unconscious structure, and a network-economic model with producer, transporter, and consumer agents decides which unconscious contents win the scarce short-term memory slots. If the topos assumption holds, consciousness becomes a formal object of study at the computational-theory level, with a precise language for talking about how perceptions, actions, and memories combine.

What carries the argument

The central object is the topos of coalgebras formed by applying a left-exact comonad to a topos, together with the Mitchell-Bénabou internal language of that topos and its Kripke-Joyal semantics. MUMBLE is simply that internal language when the topos is taken to be the ensemble of unconscious processes. The machinery works by letting logical connectives and quantifiers be arrows of the topos: conjunction and disjunction become meet and join in a Heyting algebra of subobjects, equality is read off the diagonal map, and existential and universal quantification are defined by epic covers and generalized elements. The functor from this topos into short-term memory does the 'consciousness' work, while the two transport mechanisms—Universal Reinforcement Learning for conscious-to-unconscious flow and a network economy solved by variational inequalities for the reverse flow—give the dynamics.

What would settle it

To settle the central claim, one could look for a categorical counterexample inside the proposed categories: exhibit two action-value functions whose categorical product or exponential does not exist, or show that the subobject classifier for $C_Q$ fails to be a Heyting algebra; either would destroy the topos foundation and with it the MUMBLE semantics. A behavioral falsifier would be to show that the content of short-term memory is not responsive to competitive transport costs—for example, that making a particular unconscious pathway more expensive to access leaves the probability of that content entering awareness unchanged.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that the mind's architecture can be summarized by a functor $F$ from a topos of unconscious coalgebras to conscious short-term memory, and that the internal language of that topos is a specific formal system, MUMBLE. The argument rests on two formal pillars: the earlier claim, stated as Theorem 1, that the category $C_Q$ of action-value functions forms a topos, and the standard topos-theory result, stated as Theorem 4, that a left-exact comonad on a topos produces a topos of coalgebras. In that coalgebraic topos, every unconscious process gets a logical type, truth values become elements of a Heyting algebra rather than being simply true or false, and statements about what reaches awareness are interpreted by Kripke–Joyal forcing. Conscious contents are the objects and arrows transmitted by the functor, and the competition for short-term memory is settled by a variational-inequality equilibrium of a network economy rather than by a global clock. The paper is explicit that this is a computational-theory level proposal; implementation and neural plausibility are left to future work.

Load-bearing premise

The framework assumes, without derivation or empirical support, that the ensemble of unconscious mental processes genuinely forms a topos—that the category of action-value functions has all finite limits and colimits, a subobject classifier, and exponentials, and that the comonad used to form the coalgebra topos is left exact—and it inherits the earlier claim that $C_Q$ is a topos rather than proving it here.

Editorial extensions

If this is right

  • Because every topos has an internal logic, the framework implies that the language of thought is not an arbitrary coding scheme but is forced by the category structure of unconscious processes.
  • The internal logic is intuitionistic, so the framework predicts that conscious reasoning about unconscious contents will not, in general, obey classical double-negation elimination.
  • Unconscious processing needs no global clock: both the URL consolidation and the variational-inequality competition are asynchronous and distributed, so the theory is compatible with the brain's lack of a central synchronizing signal.
  • Short-term memory contents are an equilibrium outcome of competitive bidding among unconscious processes, so which information 'wins' depends on transport costs and demand, not on a fixed winner-take-all tree.
  • The Conscious Turing Machine's binary up-tree competition is replaced by arbitrary functor diagrams, freeing the theory from a particular wiring diagram.

Reading between the lines

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

  • Editorial inference: the topos assumption could be tested piecewise—for any two unconscious processes the categorical product and exponential must exist; exhibiting a natural pair of processes that fails to compose associatively would undercut the topos foundation.
  • Editorial inference: if MUMBLE is genuinely the internal language, a discriminating experiment is to look for behavior that respects intuitionistic logic rather than classical logic, since the Kripke–Joyal semantics makes excluded middle fail in general.
  • Editorial inference: the network-economic half suggests a concrete intervention the paper does not state—raising the effective 'transport cost' between a specific long-term memory source and short-term memory should reduce that content's chance of entering awareness, which could be probed with attention or disruption experiments.
  • Editorial inference: the paper leaves the exact variance of the consciousness functor (full, faithful, or adjoint) unspecified; determining which holds would connect this proposal to the adjoint-functor integration listed as future work and would sharpen what 'preserving relationships' means for conscious contents.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

5 major / 6 minor

Summary. The paper proposes a theoretical framework called "Consciousness as a Functor" (CF), in which the ensemble of unconscious processes is modeled as a topos category of coalgebras. The internal Mitchell-Benabou language of this topos, named MUMBLE, is proposed as the "language of thought" (a categorical analogue of Brainish in the Conscious Turing Machine). The framework is positioned as a categorial formulation of Global Workspace Theory. Information flow from conscious short-term memory to unconscious long-term memory is modeled with Universal Reinforcement Learning (URL), and the reverse flow is modeled as a network economy with producer, transporter, and consumer agents, solved through asynchronous variational inequality algorithms. The paper is primarily a conceptual proposal with extensive background review of category theory, coalgebra, topos theory, and network economics.

Significance. If the topos-theoretic claims were rigorously established, CF would offer an ambitious formal unification of Global Workspace Theory with categorical logic, and MUMBLE would provide a concrete candidate for a formal internal language of thought. The paper draws on legitimate and powerful tools—topos internal logic, coalgebraic semantics, and monotone variational inequalities—and the idea of casting attentional competition as a network economy is novel. However, the central mathematical assertions are currently posited rather than proven: the topos structure of the value-function category is imported from a self-cited preprint, the comonad underlying the topos of coalgebras is never defined, and the internal language MUMBLE is simply the standard consequence of an assumed topos structure. The paper therefore does not yet deliver a validated theory, but it could serve as a research proposal from which a rigorous theory might be developed.

major comments (5)
  1. [Section 5.1, Theorem 1] Theorem 1 states that the category C_Q of value functions forms a topos, citing the author's earlier preprint [Mahadevan, 2025d]. However, the objects and morphisms of C_Q are not defined in this manuscript, and no proof or construction of limits, colimits, exponentials, and the subobject classifier is provided. This is load-bearing because the coalgebra category in Section 6.3 and the internal language MUMBLE in Section 7 rely on the topos structure of C_Q. A theorem cited from a separate, not-yet-vetted preprint cannot carry this weight without at least a statement of the precise assumptions and a proof sketch.
  2. [Section 6.3, Theorem 4] The paper quotes Mac Lane and Moerdijk's theorem that a left-exact comonad (G, ε, δ) on a topos E yields a topos E_G of coalgebras, but it never defines the endofunctor G, the counit ε, the comultiplication δ, or the ambient topos E. The central claim that "the ensemble of coalgebras defining the unconscious processes defines a topos category" (Section 3, item 2) is therefore an unproved posit. Without an explicit comonad and a verification of left-exactness, the existence of the topos of unconscious processes is not established, and everything built on it—including MUMBLE—has no foundation.
  3. [Section 1 and Section 3, item 2] The paper states that "the category of coalgebras forms a topos" as if this held for arbitrary endofunctors. This is false in general: for an arbitrary endofunctor F, the category of F-coalgebras need not have a subobject classifier or exponentials, which are essential to topos structure. The correct statement, as used in Theorem 4, requires a left-exact comonad on a topos. The sweeping claim in the Introduction is not merely imprecise; it obscures the fact that the topos property is a nontrivial condition that has not been verified for the unconscious-process coalgebras.
  4. [Section 7, MUMBLE] MUMBLE is defined as the Mitchell-Benabou internal language of the assumed topos of unconscious coalgebras. Once the topos assumption is granted, the existence of the internal language follows from standard topos theory, so the "derivation" of Brainish/MUMBLE is tautological relative to that assumption. The text says "We show formally how ... this internal language arises" (Section 1), but no formal proof is given beyond invoking textbook theorems. The paper would need to establish the topos structure independently, or explicitly treat it as an axiom and then derive empirically testable consequences, for MUMBLE to be a substantive contribution.
  5. [Sections 1–3, 'Consciousness as a Functor'] Despite the title, the paper never formally defines the alleged functor of consciousness. Section 3 discusses a "diagram functor" from the topos of unconscious processes to conscious short-term memory, but the objects, arrows, and functoriality conditions (preservation of identity and composition) are not specified. Without a precise categorical definition, "consciousness as a functor" remains a metaphor rather than a mathematical claim, which weakens the paper's central thesis.
minor comments (6)
  1. [Abstract and Section 1] The expansion of MUMBLE is inconsistent: the Abstract says "Multi-modal Universal Mitchell-Benabou Language Embedding," while Section 1 says "Multi-modal Universal Language for Mitchell-Benabou Embeddings." Please harmonize the terminology.
  2. [Section 6.1] The associativity isomorphism in Definition 10 is misprinted: "C1⊗ (C2⊗ C3) ≅ (C1⊗ C2)⊗ C2" should read "C1⊗ (C2⊗ C3) ≅ (C1⊗ C2)⊗ C3."
  3. [Section 7.3] The heading "Kripe-Joyal Semantics" is a typo for "Kripke-Joyal Semantics." Similar typos appear for "Mitchell-B'enabou" and "leke Moerdijk" throughout the text.
  4. [Section 8.3, Algorithm 2] The notation in Algorithm 2 is confusing: variables such as "Xc f" and "X f" appear without clear definitions, and the indices in the asynchronous update equations do not match the surrounding text. Please rewrite the algorithm with consistent indexing.
  5. [Section 2] There is a typo "seom from the standpoint of neuroscience" that should be "some from the standpoint." The manuscript contains many similar typographical errors and would benefit from careful proofreading.
  6. [References] Reference [Alfredo N. Iusem, 2018] is listed as "Philip Thompson Alfredo N. Iusem, Alejandro Jofré"; the author name is malformed. Also, reference [Mahadevan, 2025e] is cited as a book "In Press" and is used for several key claims; please clarify its status.

Circularity Check

2 steps flagged · score 7.0 of 10

The central 'topos of coalgebras / MUMBLE' claim rests on an unproved self-citation and on defining MUMBLE as the standard internal language of an assumed topos, so the headline 'language of thought' result is largely definitional rather than derived.

  1. self citation load bearing [Section 5 and Section 5.1, Theorem 1]
    ""We briefly summarize the result shown in our earlier paper [Mahadevan, 2025d] on URL that the category of action-value functions C_Q forms a topos, and refer the reader to that paper for details." "Theorem 1. [Mahadevan, 2025d] The category C_Q of value functions forms a topos.""

    This theorem is the only support given for the topos of value functions and, via Theorem 4, for the topos of coalgebras that is supposed to carry MUMBLE. The proof is not reproduced and the assumptions are not stated; the citation is to the author's own arXiv preprint, so the load-bearing premise is imported by self-citation rather than established here. If the cited paper contained an independent proof the citation would be legitimate, but within this manuscript the central structure rests on an unverifiable self-citation.

  2. renaming known result [Abstract and Section 7.4]
    ""As every topos has an internal language defined by a Mitchell-Bénabou language with a Kripke-Joyal semantics, the internal 'language of thought' in CF is defined as a Multi-modal Universal Mitchell-Bénabou Language Embedding (MUMBLE)." "We now have a precise semantics for MUMBLE’ing, namely the internal language of a topos, which can now serve as a 'language of thought' for our CF framework.""

    MUMBLE is not independently derived; it is defined to be the Mitchell-Bénabou internal language of the posited topos of unconscious processes. The claimed result that unconscious processes have a 'language of thought' is therefore the standard internal-language theorem applied to an assumed topos, with the target notion renamed as MUMBLE. The conclusion is contained in the definition of MUMBLE plus the assumption that the ensemble of unconscious processes forms a topos; no independent characterization of a language of thought is shown to be satisfied.

full rationale

The paper contains genuine non-circular components: the asynchronous URL and network-economy sections rely on external algorithms and results (Bertsekas-Tsitsiklis, Witsenhausen, Nagurney, Wang-Bertsekas) that are not derived from the topos claim. However, the headline contribution—that unconscious processes form a topos of coalgebras and that MUMBLE is the resulting language of thought—is not derived in this manuscript. Theorem 1 (C_Q is a topos) is imported by self-citation to [Mahadevan, 2025d] with no proof or stated assumptions, and Theorem 4 is quoted correctly but never instantiated: no comonad (G, ε, δ) on an unconscious-process topos is defined, and left-exactness is never checked. MUMBLE is then defined as the Mitchell-Bénabou internal language of that assumed topos, so the 'language of thought' conclusion is the standard internal-language construction applied to an assumed structure, i.e., a definitional reduction rather than an empirical derivation. The correct quotation of MacLane-Moerdijk is not itself circular; the circularity lies in using an unproved self-cited topos premise and then defining the target notion as its immediate consequence. This warrants a score of 7 rather than 0–2 because the central claim reduces, by the paper's own definitions and self-citation, to its assumed inputs.

Assumptions & free parameters 0 free parameters · 8 assumptions · 3 invented entities

The paper carries a heavy axiomatic load. It posits that unconscious processes form a topos of coalgebras and that value functions form a topos, with the latter resting on a self-citation. It also imports the copy/delete neural requirement and the asynchronous distributed computation paradigm as domain assumptions. No numerical free parameters are fitted, but the categorical choices (endofunctor, comonad, utility functions) are left unconstrained, which makes the framework untestable as currently stated.

assumptions (8)
  • domain assumption The ensemble of unconscious processes forms a topos category of coalgebras.
    Posited in Section 3, item 2 and Section 6.3; no derivation from neural or behavioral evidence, and the necessary left-exactness of the comonad is not shown.
  • domain assumption The category C_Q of action-value functions is a topos.
    Theorem 1, proof delegated to the author's own prior paper [Mahadevan, 2025d]; no proof is given in this text.
  • domain assumption Conscious and unconscious memory are two categories linked by functors, and consciousness is the functor.
    Central modeling choice in the Abstract and Section 3; it is an interpretive frame, not a derived result.
  • standard math Every topos has an internal Mitchell-Benabou language with Kripke-Joyal semantics.
    Used in Section 7; standard textbook result from MacLane and Moerdijk 1994.
  • standard math The category of coalgebras for a left-exact comonad on a topos is itself a topos.
    Theorem 4, MacLane and Moerdijk 1994; relied on in Section 6.3.
  • domain assumption The brain must support copy, delete, and multiple objects to perform probabilistic computation.
    Section 6.2, citing Fritz 2020; imported as a neural requirement without direct evidence in this paper.
  • domain assumption Asynchronous distributed computation without a global clock is the right computational model for unconscious processing.
    Sections 8 and 9; motivated by Bertsekas and Tsitsiklis and Witsenhausen, but the applicability to the brain is assumed.
  • standard math The two-step stochastic projection algorithm converges under Assumptions 1-6.
    Theorem 6, proof follows Iusem et al. 2018 and Wang and Bertsekas 2015.
invented entities (3)
  • MUMBLE (Multi-modal Universal Mitchell-Benabou Language Embedding)
    purpose: Defines the internal language of thought, or Brainish, in the CF framework.
    MUMBLE is the standard internal language of a topos relabeled; no empirical prediction or measurement is attached, so there is no falsifiable handle.
  • Consciousness functor (CF)
    purpose: Maps between the category of unconscious processes and conscious short-term memory.
    Defined abstractly as a functor; no specific arrows or objects are instantiated, so it cannot be tested.
  • Producer, transporter, and consumer agents in a consciousness network economy
    purpose: Model competition of unconscious processes to enter short-term memory.
    An analogy to Nagurney's network economics; no mapping to neural data, no predicted observables.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Consciousness as a Functor." pith.science (2026). https://pith.science/paper/7CGGBM3M

@misc{pith2026250817561,
  author       = {Pith},
  title        = {Pith review of: Consciousness as a Functor},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7CGGBM3M}},
  note         = {Machine review of arXiv:2508.17561}
}
read the original abstract

We propose a novel theory of consciousness as a functor (CF) that receives and transmits contents from unconscious memory into conscious memory. Our CF framework can be seen as a categorial formulation of the Global Workspace Theory proposed by Baars. CF models the ensemble of unconscious processes as a topos category of coalgebras. The internal language of thought in CF is defined as a Multi-modal Universal Mitchell-Benabou Language Embedding (MUMBLE). We model the transmission of information from conscious short-term working memory to long-term unconscious memory using our recently proposed Universal Reinforcement Learning (URL) framework. To model the transmission of information from unconscious long-term memory into resource-constrained short-term memory, we propose a network economic model.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

81 extracted references · 37 canonical work pages

  1. [1]

    Non-Well-Founded Sets

    Peter Aczel. Non-Well-Founded Sets. CSLI Lecture Notes, Palo Alto, CA, USA, 1988

    1988

  2. [2]

    A final coalgebra theorem

    Peter Aczel and Nax Mendler. A final coalgebra theorem. In David H. Pitt, David E. Rydeheard, Peter Dybjer, Andrew M. Pitts, and Axel Poign \'e , editors, Category Theory and Computer Science, pages 357--365, Berlin, Heidelberg, 1989. Springer Berlin Heidelberg. ISBN 978-3-540-46740-3

    work page 1989

  3. [3]

    Iusem, Alejandro Jofré

    Philip Thompson Alfredo N. Iusem, Alejandro Jofré. Incremental constraint projection methods for monotone stochastic variational inequalities. Mathematics of Operations Research, 44 0 (1): 0 236--263, 2018. URL https://doi.org/10.1287/moor.2017.0922

    work page arXiv 2018

  4. [4]

    Arbib and Ernest G

    Michael A. Arbib and Ernest G. Manes. A categorist's view of automata and systems. In Ernest G. Manes, editor, Category Theory Applied to Computation and Control, Proceedings of the First International Symposium, San Francisco, CA, USA, February 25-26, 1974, Proceedings, volume 25 of Lecture Notes in Computer Science, pages 51--64. Springer, 1974. doi:10....

    work page doi:10.1007/3-540-07142-3 1974

  5. [5]

    Computational Complexity - A Modern Approach

    Sanjeev Arora and Boaz Barak. Computational Complexity - A Modern Approach . Cambridge University Press, 2009. ISBN 978-0-521-42426-4. URL http://www.cambridge.org/catalogue/catalogue.asp?isbn=9780521424264

    work page 2009

  6. [6]

    In the theater of consciousness: The workspace of the mind

    Bernard Baars. In the theater of consciousness: The workspace of the mind. Oxford Univ. Press, New York, NY, 1997 a . URL https://psycnet.apa.org/doi/10.1093/acprof:oso/9780195102659.001.1

    arXiv 1997

  7. [7]

    In the Theater of Consciousness: The Workspace of the Mind

    Bernard Baars. In the Theater of Consciousness: The Workspace of the Mind. Oxford University Press, 1997 b

    work page 1997

  8. [8]

    J. L. Bell. Toposes and Local Set Theories. Dover, 1988

    1988

Show all 81 references
  1. [9]

    The consciousness prior, 2019

    Yoshua Bengio. The consciousness prior, 2019. URL https://arxiv.org/abs/1709.08568

    2019 arXiv

  2. [10]

    Reinforcement Learning and Optimal Control

    Dimitri Bertsekas. Reinforcement Learning and Optimal Control. Athena Scientific, 2019

    2019

  3. [11]

    Bertsekas

    Dimitri P. Bertsekas. Dynamic programming and optimal control, 3rd Edition. Athena Scientific, 2005. ISBN 1886529264. URL https://www.worldcat.org/oclc/314894080

    2005

  4. [12]

    Bertsekas and John N

    Dimitri P. Bertsekas and John N. Tsitsiklis. Parallel and Distributed Computation: Numerical Methods. Athena Scientific, 1997. ISBN 1886529019

    1997

  5. [13]

    A theory of consciousness from a theoretical computer science perspective: Insights from the conscious turing machine

    Lenore Blum and Manuel Blum. A theory of consciousness from a theoretical computer science perspective: Insights from the conscious turing machine. Proceedings of the National Academy of Sciences, 119 0 (21): 0 e2115934119, 2022. doi:10.1073/pnas.2115934119. URL https://www.pn...

    2022 doi

  6. [14]

    A theoretical computer science perspective on consciousness, 2021

    Manuel Blum and Lenore Blum. A theoretical computer science perspective on consciousness, 2021. URL https://arxiv.org/abs/2011.09850

    2021 arXiv

  7. [15]

    Bradley, J

    TD. Bradley, J. Terilla, and Y. Vlassopoulos. An enriched category theory of language: From syntax to semantics. La Matematica, 1: 0 551--580, 2022

    2022

  8. [16]

    Classifying clustering schemes, 2010

    Gunnar Carlsson and Facundo Memoli. Classifying clustering schemes, 2010. URL http://arxiv.org/abs/1011.5270. cite arxiv:1011.5270

    2010 arXiv

  9. [17]

    Carpentier, J.P

    P. Carpentier, J.P. Chancelier, G. Cohen, and M. De Lara. Stochastic Multi-Stage Optimization: At the Crossroads Between Discrete Time Stochastic Control and Stochastic Programming. Probability theory and stochastic modelling. Springer International Publishing, 2015. ISBN 9783...

    2015

  10. [18]

    Chalmers

    D. Chalmers. The Conscious Mind. Oxford University Press, 1996 a

    1996

  11. [19]

    The Conscious Mind: In Search of a Fundamental Theory

    David Chalmers. The Conscious Mind: In Search of a Fundamental Theory. Oxford, 1996 b

    1996

  12. [20]

    Disintegration and bayesian inversion via string diagrams

    Kenta Cho and Bart Jacobs. Disintegration and bayesian inversion via string diagrams. Mathematical Structures in Computer Science, 29 0 (7): 0 938–971, March 2019. ISSN 1469-8072. doi:10.1017/s0960129518000488. URL http://dx.doi.org/10.1017/S0960129518000488

    2019 doi

  13. [21]

    On certain formal properties of grammars

    Noam Chomsky. On certain formal properties of grammars. Information and Control, 2 0 (2): 0 137--167, June 1959. URL http://www.diku.dk/hjemmesider/ansatte/henglein/papers/chomsky1959.pdf

    1959

  14. [22]

    Aspects of the Theory of Syntax

    Noam Chomsky. Aspects of the Theory of Syntax. The MIT Press, Cambridge, 1965. URL http://www.amazon.com/Aspects-Theory-Syntax-Noam-Chomsky/dp/0262530074

    1965

  15. [23]

    Toward a neurobiological theory of consciousness

    Francis Crick and Christof Koch. Toward a neurobiological theory of consciousness. Seminars in the Neurosciences, 2: 0 263--275, 1990

    1990

  16. [24]

    Cpnsciousness and the Brain

    Stanis;as Dehaene. Cpnsciousness and the Brain. Penguin Books, 2014

    2014

  17. [25]

    D. C. Dennett. Consciousness Explained. Little, Brown, and Company, 1991

    1991

  18. [26]

    Principles of Philosophy

    Rene Descartes. Principles of Philosophy. Dordecht: Reidel, 1644

  19. [27]

    A Universe of Consciousness: How Matter becomes Imagination

    Gerard Edelman and Giulio Tononi. A Universe of Consciousness: How Matter becomes Imagination. Basic Books, 2000

    2000

  20. [28]

    Facchinei and J

    F. Facchinei and J. Pang. Finite-Dimensional Variational Inequalities and Complementarity Problems. Springer, 2003

    2003

  21. [29]

    Frank M. V. Feys, Helle Hvid Hansen, and Lawrence S. Moss. Long-term values in markov decision processes, (co)algebraically. In Coalgebraic Methods in Computer Science: 14th IFIP WG 1.3 International Workshop, CMCS 2018, Colocated with ETAPS 2018, Thessaloniki, Greece, April 1...

    2018 doi

  22. [30]

    The Language of Thought

    Jerry Fodor. The Language of Thought. Harvard University Press, USA, 1980. ISBN 9780674510302

    1980

  23. [31]

    Seven Sketches in Compositionality: An Invitation to Applied Category Theory

    Brendan Fong and David I Spivak. Seven Sketches in Compositionality: An Invitation to Applied Category Theory. Cambridge University Press, 2018

    2018

  24. [32]

    Universal Coalgebras

    James Fox. Universal Coalgebras. Dissertation, McGill University, 1976

    1976

  25. [33]

    A synthetic approach to markov kernels, conditional independence and theorems on sufficient statistics

    Tobias Fritz. A synthetic approach to markov kernels, conditional independence and theorems on sufficient statistics. Advances in Mathematics, 370: 0 107239, August 2020. ISSN 0001-8708. doi:10.1016/j.aim.2020.107239. URL http://dx.doi.org/10.1016/j.aim.2020.107239

    2020

  26. [34]

    The d-separation criterion in categorical probability

    Tobias Fritz and Andreas Klingler. The d-separation criterion in categorical probability. Journal of Machine Learning Research, 24 0 (46): 0 1--49, 2023. URL http://jmlr.org/papers/v24/22-0916.html

    2023

  27. [35]

    Gallistel and Adam Philip King

    C.R. Gallistel and Adam Philip King. Memory and the Computational Brain. Wiley-Blackwell, United States, 2010

    2010

  28. [36]

    Topoi: The Categorial Analysis of Logic

    Robert Goldblatt. Topoi: The Categorial Analysis of Logic. Dover Press, 2006

    2006

  29. [37]

    Categories for Quantum Theory: An Introduction

    Chris Heunen and Jamie Vicary. Categories for Quantum Theory: An Introduction. Oxford University Press, November 2019. doi:10.1093/oso/9780198739623.001.0001

    2019

  30. [38]

    Introduction to Coalgebra: Towards Mathematics of States and Observation, volume 59 of Cambridge Tracts in Theoretical Computer Science

    Bart Jacobs. Introduction to Coalgebra: Towards Mathematics of States and Observation, volume 59 of Cambridge Tracts in Theoretical Computer Science. Cambridge University Press, 2016. ISBN 9781316823187. doi:10.1017/CBO9781316823187. URL https://doi.org/10.1017/CBO9781316823187

    2016 doi

  31. [39]

    Causal inference via string diagram surgery: A diagrammatic approach to interventions and counterfactuals

    Bart Jacobs, Aleks Kissinger, and Fabio Zanasi. Causal inference via string diagram surgery: A diagrammatic approach to interventions and counterfactuals. Math. Struct. Comput. Sci., 31 0 (5): 0 553--574, 2021. doi:10.1017/S096012952100027X. URL https://doi.org/10.1017/S096012...

    2021 doi

  32. [40]

    Sketches of an elephant: a Topos theory compendium

    Peter T Johnstone. Sketches of an elephant: a Topos theory compendium . Oxford logic guides. Oxford Univ. Press, New York, NY, 2002. URL https://cds.cern.ch/record/592033

    2002

  33. [41]

    Topos Theory

    Peter T Johnstone. Topos Theory . Dover Publications, 2014

    2014

  34. [42]

    Thinking, fast and slow

    Daniel Kahneman. Thinking, fast and slow. Farrar, Straus and Giroux, New York, 2011. ISBN 9780374275631 0374275637. URL https://www.amazon.de/Thinking-Fast-Slow-Daniel-Kahneman/dp/0374275637/ref=wl_it_dp_o_pdT1_nS_nC?ie=UTF8&colid=151193SNGKJT9&coliid=I3OCESLZCVDFL7

    2011

  35. [43]

    Applications of metric coinduction

    Dexter Kozen and Nicholas Ruozzi. Applications of metric coinduction. Log. Methods Comput. Sci., 5 0 (3), 2009. URL http://arxiv.org/abs/0908.2793

    2009 arXiv

  36. [44]

    J. Locke. John Locke: An Essay concerning Human Understanding. Oxford World's Classics. Oxford University Press, United States, 1688. ISBN 978-0-19-929662-0

  37. [45]

    Categories for the Working Mathematician

    Saunders MacLane. Categories for the Working Mathematician. Springer-Verlag, New York, 1971. Graduate Texts in Mathematics, Vol. 5

    1971

  38. [46]

    Sheaves in Geometry and Logic: A First Introduction to Topos Theory

    Saunders MacLane and leke Moerdijk. Sheaves in Geometry and Logic: A First Introduction to Topos Theory. Springer, 1994

    1994

  39. [48]

    Universal imitation games

    Sridhar Mahadevan. Universal imitation games. CoRR, abs/2110.15431, 2021 b . URL https://arxiv.org/abs/2110.15431

    2021 arXiv

  40. [49]

    Universal causality

    Sridhar Mahadevan. Universal causality. Entropy, 25 0 (4): 0 574, 2023. doi:10.3390/E25040574. URL https://doi.org/10.3390/e25040574

    2023 doi

  41. [50]

    Gaia: Categorical foundations of generative ai, 2024

    Sridhar Mahadevan. Gaia: Categorical foundations of generative ai, 2024. URL https://arxiv.org/abs/2402.18732

    2024 arXiv

  42. [51]

    A higher algebraic k-theory of causality

    Sridhar Mahadevan. A higher algebraic k-theory of causality. Entropy, In Press 2025 a . doi:10.20944/preprints202501.1242.v1. URL https://doi.org/10.20944/preprints202501.1242.v1

    2025

  43. [52]

    Topos causal models, 2025 b

    Sridhar Mahadevan. Topos causal models, 2025 b . URL https://arxiv.org/abs/2508.08295

    2025 arXiv

  44. [53]

    Topos theory for generative ai and llms, 2025 c

    Sridhar Mahadevan. Topos theory for generative ai and llms, 2025 c . URL https://arxiv.org/abs/2508.08293

    2025 arXiv

  45. [54]

    Universal reinforcement learning in coalgebras: Asynchronous stochastic computation via conduction, 2025 d

    Sridhar Mahadevan. Universal reinforcement learning in coalgebras: Asynchronous stochastic computation via conduction, 2025 d . URL https://arxiv.org/abs/2508.15128

    2025 arXiv

  46. [55]

    Artificial General Intelligence: A Categorial Theory

    Sridhar Mahadevan. Artificial General Intelligence: A Categorial Theory. Springer, 2025 e . In Press

    2025

  47. [56]

    Vision: A Computational Investigation into the Human Representation and Processing of Visual Information

    David Marr. Vision: A Computational Investigation into the Human Representation and Processing of Visual Information. Henry Holt and Co., Inc., New York, NY, USA, 1982. ISBN 0716715678

    1982

  48. [57]

    AI: a personal view, page 97–107

    David Marr. AI: a personal view, page 97–107. Cambridge University Press, USA, 1990. ISBN 0521359449

    1990

  49. [58]

    Game Theory

    Michael Maschler, Eilon Solan, and Shmuel Zamir. Game Theory. Cambridge University Press, 2013

    2013

  50. [59]

    Umap: Uniform manifold approximation and projection for dimension reduction, 2018

    Leland McInnes, John Healy, and James Melville. Umap: Uniform manifold approximation and projection for dimension reduction, 2018. URL https://arxiv.org/abs/1802.03426

    2018 arXiv

  51. [60]

    Boolean topoi and the theory of sets

    William Mitchell. Boolean topoi and the theory of sets. LJournal of Pure and Applied Algebra, 2: 0 261--274, 1972

    1972

  52. [61]

    Levels of AGI

    Meredith Ringel Morris, Jascha Sohl-Dickstein, Noah Fiedel, Tris Warkentin, Allan Dafoe, Aleksandra Faust, Clement Farabet, and Shane Legg, editors. Levels of AGI for Operationalizing Progress on the Path to AGI, 2023. Original arXiv title in November 2023 was "Levels of AGI":...

    2023

  53. [62]

    Nagurney

    A. Nagurney. Network Economics: A Variational Inequality Approach. Kluwer Academic Press, 1999

    1999

  54. [63]

    Non-cooperative games

    John Nash. Non-cooperative games. Annals of Mathematics, 54 0 (2): 0 286--295, 1951. URL https://doi.org/10.2307/1969529

    1951 doi

  55. [64]

    Algorithmic Game Theory

    Noam Nisan, Tim Roughgarden, Eva Tardos, and Vijay Vazirani. Algorithmic Game Theory. Cambridge University Press, Cambridge; New York, 2007. ISBN 9780521872829 0521872820

    2007

  56. [65]

    Causality: Models, Reasoning and Inference

    Judea Pearl. Causality: Models, Reasoning and Inference. Cambridge University Press, USA, 2nd edition, 2009. ISBN 052189560X

    2009

  57. [66]

    Puterman

    Martin L. Puterman. Chapter 8 markov decision processes. In Stochastic Models, volume 2 of Handbooks in Operations Research and Management Science, pages 331--434. Elsevier, 1990. doi:https://doi.org/10.1016/S0927-0507(05)80172-0. URL https://www.sciencedirect.com/science/arti...

    1990 doi

  58. [67]

    B. Richter. From Categories to Homotopy Theory. Cambridge Studies in Advanced Mathematics. Cambridge University Press, 2020. ISBN 9781108479622. URL https://books.google.com/books?id=pnzUDwAAQBAJ

    2020

  59. [68]

    E. Riehl. Category Theory in Context. Aurora: Dover Modern Math Originals. Dover Publications, 2017. ISBN 9780486820804. URL https://books.google.com/books?id=6B9MDgAAQBAJ

    2017

  60. [69]

    A Stochastic Approximation Method

    Herbert Robbins and Sutton Monro. A Stochastic Approximation Method . The Annals of Mathematical Statistics, 22 0 (3): 0 400 -- 407, 1951. doi:10.1214/aoms/1177729586. URL https://doi.org/10.1214/aoms/1177729586

    1951

  61. [70]

    J. J. M. M. Rutten. A coinductive calculus of streams. Mathematical. Structures in Comp. Sci., 15 0 (1): 0 93–147, February 2005. ISSN 0960-1295. doi:10.1017/S0960129504004517. URL https://doi.org/10.1017/S0960129504004517

    2005 doi

  62. [71]

    J.J.M.M. Rutten. Universal coalgebra: a theory of systems. Theoretical Computer Science, 249 0 (1): 0 3 -- 80, 2000. ISSN 0304-3975. doi:http://dx.doi.org/10.1016/S0304-3975(00)00056-6. URL http://www.sciencedirect.com/science/article/pii/S0304397500000566. Modern Algebra

    2000 doi

  63. [72]

    John R. Searle. The Mystery of Consciousness. New York Review Book, 1997

    1997

  64. [73]

    Selinger

    P. Selinger. A survey of graphical languages for monoidal categories. In New Structures for Physics, pages 289--355. Springer Berlin Heidelberg, 2010. doi:10.1007/978-3-642-12821-9_4. URL https://doi.org/10.1007

    2010 doi

  65. [74]

    Singh, Michael R

    Satinder P. Singh, Michael R. James, and Matthew R. Rudary. Predictive state representations: A new theory for modeling dynamical systems. In David Maxwell Chickering and Joseph Y. Halpern, editors, UAI '04, Proceedings of the 20th Conference in Uncertainty in Artificial Intel...

    2004

  66. [75]

    Probabilistic systems coalgebraically: A survey

    Ana Sokolova. Probabilistic systems coalgebraically: A survey. Theoretical Computer Science, 412 0 (38): 0 5095--5110, 2011. ISSN 0304-3975. doi:https://doi.org/10.1016/j.tcs.2011.05.008. URL https://www.sciencedirect.com/science/article/pii/S0304397511003902. CMCS Tenth Anniv...

    2011 doi

  67. [76]

    Spivak and Robert E

    David I. Spivak and Robert E. Kent. Ologs: A categorical framework for knowledge representation. PLoS ONE , 7 0 (1): 0 e24274, jan 2012. doi:10.1371/journal.pone.0024274

    2012 doi

  68. [77]

    Sutton and Andrew G

    Richard S. Sutton and Andrew G. Barto. Reinforcement Learning: An Introduction. The MIT Press, second edition, 2018. URL http://incompleteideas.net/book/the-book-2nd.html

    2018

  69. [78]

    Tsitsiklis

    J.N. Tsitsiklis. Asynchronous stochastic approximation and q-learning. In Proceedings of 32nd IEEE Conference on Decision and Control, pages 395--400 vol.1, 1993. doi:10.1109/CDC.1993.325119

    1993

  70. [79]

    A guided tour in the topos of graphs, 2003

    Sebastiano Vigna. A guided tour in the topos of graphs, 2003

    2003

  71. [80]

    von Neumann and O

    J. von Neumann and O. Morgenstern. Theory of games and economic behavior. Princeton University Press, 1947

    1947

  72. [81]

    Bertsekas

    Mengdi Wang and Dimitri P. Bertsekas. Incremental constraint projection methods for variational inequalities. Math. Program., 150 0 (2): 0 321--363, 2015. doi:10.1007/s10107-014-0769-x. URL https://doi.org/10.1007/s10107-014-0769-x

    2015 doi

  73. [82]

    H. S. Witsenhausen. The intrinsic model for discrete stochastic control: Some open problems. In A. Bensoussan and J. L. Lions, editors, Control Theory, Numerical Methods and Computer Systems Modelling, pages 322--335, Berlin, Heidelberg, 1975. Springer Berlin Heidelberg

    1975

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.