A Computational Model of Learning and Memory Using Structurally Dynamic Cellular Automata

Jeet Singh

REVIEW 5 major objections 7 minor 36 references

A Computational Model of Learning and Memory Using Structurally Dynamic Cellular Automata

T0 review · 5 major / 7 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Coincidence detection, signal modulation, and reward/penalty edges suffice to create an information space that lets a graph agent learn, remember, and navigate to a reward in one training pass.

desk verdict A transparent toy model with a genuinely new combination of mechanisms, but the headline 'near-optimal' claim is unquantified and the evidence is too thin to support it as stated. read the letter →

arxiv 2501.06192 v1 pith:7VRWSDAI submitted 2024-12-20 cs.AI cs.NEmath.DSq-bio.NC

Jeet Singh This is my paper · ORCID

classification cs.AIcs.NEmath.DSq-bio.NC

keywords structurallydynamiccellularautomatacoincidencedetectionsignalmodulationrewardgradientone-shotlearninggraphnavigationcontext-dependentmemoryde-inforcementplasticity

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

This paper tries to establish that three minimal, bio-plausible operations—coincidence detection, signal modulation, and reward/penalty edge weighting—are sufficient to create an information space over which a simple agent can compute, remember, and choose. The agent runs a structurally dynamic cellular automaton on a graph: cell values spread through a lazy-random-walk update, edge conductances encode experience, and pre-wired reward and penalty edges tilt the resulting values into a gradient. In small lattice and Watts-Strogatz environments, one training pass lets the agent navigate to a reward, avoid penalties, switch between rewards by context, and explore under sparse reward. If the claim holds, the model offers a candidate explanation for how learning and memory could live in the structure of connections rather than in stored data.

What carries the argument

The load-bearing mechanism is a continuous-valued, structurally dynamic cellular automaton, called Coincident Graph Learning (CGL), whose cell values evolve by $\phi(x) = \tfrac12 x(D^{-1} C + I)$, the transition operator of a lazy random walk; repeated application for $r$ steps spreads an input's value through the graph and creates a graded field. The graph update $\psi$ writes memory: it connects nodes whose inputs coincide, sets new edge conductance to 1, and decrements repeated coincidences by factor $d$ down to a floor $b$. Reward and penalty edges are pre-wired conductances above one or between zero and one, so they act as sources and sinks in the propagated field. The agent's choice rule, move to the highest-value neighbor, turns that field into behavior. The recursion depth $r$ fixes the spatial range of the gradient and therefore the reach of memory and inference.

What would settle it

Pick any trained environment from the paper and any start node within the recursion range of the reward; compute the output vector $\omega(x) = \phi^r(x)$ at that node and check whether the highest-value neighbor strictly decreases the graph distance to the reward. A single node where the best-valued neighbor moves away from the reward would show that the propagated gradient is not, by itself, the mechanism producing the reported near-optimal navigation.

Watch

Extended reading notes

Core claim

The central claim is that the combination of (1) a cell update function $\phi(x) = \tfrac12 x(D^{-1}C + I)$ that behaves like a lazy random walk, (2) a recursion $\omega(x_n)=\phi^r(x_n)$ that propagates values for a fixed depth, and (3) a graph update $\psi$ that adds an edge when two inputs are coincident and de-inforces it by a factor $d$ when the coincidence repeats, is enough to make reward and penalty locations cast a usable signal gradient across the graph. Memory of a training run is not stored as data; it is embedded in the conductance pattern of the coincident graph. On a single pass through the environment, the agent builds this pattern, and afterward the highest-value neighboring node in the propagated field points toward the reward. The paper reports 92–100% success in 400 random start–target trials on 100-node graphs, median path lengths of 8–19 steps, and demonstrations of penalty avoidance, context-dependent reward choice, and exploratory behavior in sparse-reward settings.

Load-bearing premise

The central claim depends on the recursion depth $r$ being matched to the graph's diameter and connectivity; the paper changes $r$ from 6 to 3 for more random Watts-Strogatz graphs because too much recursion obscures the reward signal, and it gives no automatic rule for choosing $r$.

Editorial extensions

If this is right

A single training pass can set up a usable reward gradient in small lattices and small-world graphs, with success rates of 92–100% across 400 random start–target trials.
Memory can be regenerated on demand from the current input and the edge conductances of the coincident graph, so no separate storage of input or output sequences is required.
Context dependence and task switching are direct effects of the input vector: with two rewards pre-wired, different start contexts navigate to different rewards without retraining.
De-inforcement as the plasticity rule yields exploratory or novelty-seeking behavior when the gradient is flat, and penalizing by lowering edge conductance naturally creates avoidance.
The low cost of single-pass training and local updates suggests the model could scale to larger environments, with the recursion depth acting as the abstraction depth of the computation.

Reading between the lines

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

An inference beyond the paper: choosing the recursion depth $r$ from the graph's local diameter or from the entropy of the propagated field would remove the paper's only reported hand-tuned parameter; the paper does not propose such an adaptive rule.
A testable extension: because de-inforcement only reduces repeated coincident edges, a moved or deleted reward should leave a residual conductance trace, so the model should show proactive interference unless the old reward edge is explicitly reset; the paper does not report this forgetting experiment.
If memory truly lives in edge conductances, perturbing the coincident matrix with small noise should degrade recall proportionally to the perturbation; this experiment would distinguish CGL from a method that stores explicit trajectories.

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.

Desk Editor's Note

read the letter

Quick take: the CGL model is a clear, simple combination of coincidence-based edge creation, de-inforcement, and a lazy-random-walk cell update with pre-wired reward/penalty edges. That particular package is new, and the paper explains it with checkable equations and a worked example. The memory vector m = ω(x) − x is a neat formalization of memory-as-regeneration, and the de-inforcement mechanism as boredom-driven exploration is an interesting idea.

The soft spots are real and they are in the experimental section. 'Near-optimal' is never defined. Section 3.7 defines success as reaching the target within N steps and staying there for 3 steps, which is a very generous budget. In the 100-node lattice trial the median path length is 19 and the mean is 20.45; the expected Manhattan distance between random pairs in a 10x10 grid is about 6.7, so the reported paths are roughly three times optimal. That is not 'near-optimal' by any natural reading. The paper also gives no comparison to a no-learning baseline: if the same pre-wired reward edges and greedy policy would navigate successfully without the training run, the 'single training run' claim is empty. That baseline is missing.

The recursion depth r is acknowledged in Section 5 to need hand-tuning (from 6 to 3) as graph randomness changes; the paper offers no rule for choosing r. This is an admitted limitation, not a hidden one, but it undermines the sufficiency claim.

On the positive side, the equations are simple enough to check, the example in Section 3.3 is consistent, and the randomized trials at least report success rates and path length distributions, which is more than many toy-model papers do. Still, without code, data, baselines, or an optimality comparison, the central claims of 'near-optimal choice' and sufficiency cannot be accepted. The paper is honest about its limitations in the discussion, but the abstract overstates.

Who is this for? Researchers working on biologically-plausible learning, graph cellular automata, or energy-efficient continual learning might find the mechanism worth exploring. It deserves a serious referee, not a desk reject, because the combination is new and the formal core is transparent. But the referee should insist on code/data, a shortest-path comparison, and a no-learning control before the claims can be accepted.

Referee Report

5 major / 7 minor

Summary. This paper proposes Coincident Graph Learning (CGL), a structurally dynamic cellular automaton model in which a continuous-valued cell update function (Eq. 1) propagates activity over a graph whose edge weights are modified by a de-inforcement rule (Eq. 6), with reward/penalty circuits pre-wired into the graph. The model is tested on small lattice and Watts-Strogatz graphs; after a single training pass, the agent greedily moves toward the highest-value neighbor. The reported experiments show high success rates in reaching reward nodes within N steps and path-length distributions for 400 random start-target pairs. The paper claims these three building blocks suffice for near-optimal navigation, penalty avoidance, exploration, context-dependent memory, and computational efficiency.

Significance. If the claims were fully substantiated, the model would be a notable minimal demonstration that a lazy-random-walk propagator plus edge plasticity and pre-wired reward/penalty conductances can support one-pass navigation and memory regeneration in small graphs, with potential relevance to neuroscience-inspired AI and reservoir computing. Strengths include a compact closed-form update, a consistent worked example in Section 3.3, and a nontrivial randomized evaluation with 400 pairs per topology. However, the significance is currently limited because the headline 'near-optimal' claim is not quantified, no baselines are reported, and key parameters and training details are underspecified or hand-tuned.

major comments (5)

[Abstract, §3.7, §4.6] The claim that the model 'can make near-optimal choices' is not supported by the success metric. Section 3.7 defines success only as reaching the reward within N steps and remaining there for at least 3 steps, and Section 4.6 reports a mean path length of 20.45 and median 19 for the 100-node lattice trial; for random start-target pairs on a 10x10 grid the mean Manhattan shortest path is approximately 6.7, so the reported paths are roughly 2-3 times optimal. Please define the optimality metric, report the distribution of excess path length over the shortest path for each of the 400 pairs, and include a no-learning baseline (e.g., random walk or greedy gradient on the raw environment adjacency matrix without training) to demonstrate that the single training run materially improves navigation.
[§4.6, §5] The recursion range r is a hand-tuned parameter that is changed between trials (r=6 for the lattice and 0% WS graphs, r=3 for 50% and 100% WS graphs), and the Discussion explicitly states that too much recursion 'generates noise in the gradient and obscures the signal.' No rule is given for selecting r from graph statistics, and no sensitivity analysis is reported. Because the reported success depends on matching r to the environment's diameter and connectivity, the sufficiency claim for the three building blocks is not established; please provide an automatic selection criterion or show that performance is robust across a range of r for each topology.
[§3.7, §4.2] The training protocol is underspecified. The text says the agent performs 'a single complete pass through the environment' and perceives inputs directly adjacent to its current node, but it does not state whether this pass visits every node, whether all incident edges are observed from each visited node, or whether any edge is observed more than once (which is required for de-inforcement in Eq. 6 to act during training). If the pass is a full tour that reveals every node's neighbors, then the coincident graph essentially reproduces the environment adjacency matrix and the one-pass result is less surprising; please specify the exact training walk, the number of times each edge is observed, and the state of the coincident graph after training.
[§2.1, §3.5, §3.7] Reward and penalty circuits are pre-wired into the coincident graph before training, with the reward implemented as an edge of elevated conductance to a specialized reward node connected to the target. This means the agent is told the reward location through the graph construction rather than having to discover it from reward signals during interaction. To support the claim that the model learns to re-discover a reward state after a single training run, please test a condition in which the reward identity must be inferred from reward inputs experienced during training, or at least report a control in which the target is not pre-wired and show that training alone creates the gradient.
[§3.6, §4.4] The memory claims are not experimentally validated. The memory vector m is defined in Eq. (7) but the text states that 'we do not need to use the value of m directly in the model,' and no experiment measures whether m regenerates a previous experience. The context-dependent result in Section 4.4 varies start locations rather than testing reconstruction of stored representations; please add a quantitative memory retrieval task (e.g., decode the remembered target from m or from the graph state) or explicitly restrict the claims to navigation rather than memory.

minor comments (7)

[§3.4, Eq. (6)] The condition 'i ≠ 0' should be 'i ≠ j', and the piecewise rule should state explicitly that the 'otherwise' case includes C_ij ≤ b and the case of no simultaneous input.
[Figures] The figures are not all referenced in numerical order (e.g., Figure 2 is not called out in the text); please add explicit references to each figure at the point where it is discussed.
[Throughout] The paper would benefit from a careful proofreading pass; examples include 'an agent and' in Section 2.2, 'the the node values' in the Figure 3 caption, and 'an broad range' in Section 4.6.
[§4.6] Please report the Watts-Strogatz parameters (degree and rewiring probability) and the average diameter of each graph, since the choice of r is justified by diameter in Section 5.
[§3.3] In the worked example, the diagonal matrix in Eqs. (4)-(5) is the inverse degree matrix D^{-1}; this should be stated explicitly in the text.
[Abstract, §3.7] The term 'near-optimal' is used in the abstract but never defined; if it is retained, define it and use it consistently with the reported path lengths.
[Throughout] No code or data availability statement is included; providing the MATLAB implementation would improve reproducibility of the reported experiments.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the behavioral results are direct simulations of the paper's explicit update equations with disclosed reward/recursion inputs.

full rationale

The paper's claimed derivation chain runs from explicit update rules (Eq. 1 lazy random walk, Eq. 6 de-inforcement, Eq. 2 recursion) to simulated navigation; the reward/penalty inputs are explicitly pre-wired (Sec. 3.5), and the behavioral policy is explicitly defined as moving to the highest-value adjacent node (Sec. 3.7). Success rates in Figs. 8-11 are outcomes of simulating those rules, not quantities recovered from the same output they are asked to predict. The hand-tuning of recursion depth r and reward level is acknowledged in Sec. 4.6 and Sec. 5; that is model selection on test environments, a validity limitation, not a circular reduction. The memory vector m (Eq. 7) is explicitly unused ('we do not need to use the value of m directly in the model'), so it cannot be load-bearing. No self-citations are load-bearing: the cited Singh et al. 2009 is a different author's RL theory reference, and no uniqueness theorem is imported. The missing optimality baseline and undefined 'near-optimal' are evidentiary gaps, not circularity under the defined patterns.

Assumptions & free parameters 6 free parameters · 6 assumptions · 3 invented entities

The central claim rests on a small set of hand-chosen numerical parameters (d, r, reward and penalty conductances) and on assumptions about perception, coincidence windows, and pre-wired reward circuits. The invented entities are internal to the model and have no independent evidence outside the simulations.

free parameters (6)

de-inforcement factor d = 0.5
Set in Section 3.4 and used in Eq. 6; controls how quickly re-seen coincident edges weaken, which is the mechanism claimed to generate exploration.
minimum edge floor b = not specified
Lower bound below which de-inforcement stops in Eq. 6; no numerical value is given anywhere, so exact edge dynamics cannot be replicated.
recursion range r = 3 or 6 depending on graph class
Number of times the cell update is applied. Manually changed from 6 to 3 for randomized Watts-Strogatz graphs, as admitted in Section 5.
reward edge conductance = 3 or 5 across trials
Pre-wired reward edge weights in the coincident matrix; different reward levels are used in the four randomized trials (Figs. 8 to 11).
penalty edge conductance = between 0 and 1, exact values not reported
Penalty edges are coded as edge values below 1, but the specific strengths in the penalty experiments are not listed.
latency coefficient l = 0
Latent carry-over term in Eq. 3 is set to zero in all simulations, so the recursion is just repeated application of the cell update.

assumptions (6)

domain assumption A short coincidence time window lets distinct sensory events be treated as simultaneous.
Section 3 states that coincident events are sensory events within a very short time interval; this is what triggers Eq. 6 edge creation.
domain assumption The agent perceives only inputs from nodes adjacent to its current node and has no coordinates.
Section 3.7 specifies this perceptual restriction; both training and navigation depend on local adjacency information.
ad hoc to paper Reward and penalty circuits are pre-wired into the coincident matrix before training.
Section 3.5 and 3.7 treat reward/penalty nodes as intrinsic to the matrix rather than learned, which is what creates the gradient the agent follows.
standard math Repeated application of the lazy random walk update Eq. 1 produces a useful gradient over the graph.
The model relies on diffusive spread of node values but does not derive conditions under which the gradient reliably points to the reward.
ad hoc to paper De-inforcement, rather than reinforcement, is the primary plasticity rule and yields exploration.
Section 3.4 selects de-inforcement with factor d=0.5 and Section 5 argues it produces novelty-seeking; no independent biological or quantitative validation is provided.
ad hoc to paper Recursion range r can be interpreted as the abstraction capacity or logical depth of the computation.
Section 5 proposes this interpretation as a conjecture, but it is not derived or measured.

invented entities (3)

Coincident Graph
purpose: Internal memory and computation substrate: an adjacency matrix whose edge weights encode learned coincidences and conductances.
The central construct of the CGL model. Its existence is asserted and demonstrated only in the paper's simulations, with no external validation.
Reward and penalty nodes and edges
purpose: Pre-wired high- or low-conductance connections that generate the signal gradient toward rewards and away from penalties.
Inserted by hand into the coincident matrix (Section 3.5), not discovered by the agent, and their specifications are under-reported.
Memory vector m
purpose: Defines memory as the difference between reconstructed output and current input (Eq. 7).
Purely definitional; the paper says m is not used in the model, so it carries no measured or falsifiable content.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A Computational Model of Learning and Memory Using Structurally Dynamic Cellular Automata." pith.science (2026). https://pith.science/paper/7VRWSDAI

@misc{pith2026250106192,
  author       = {Pith},
  title        = {Pith review of: A Computational Model of Learning and Memory Using Structurally Dynamic Cellular Automata},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7VRWSDAI}},
  note         = {Machine review of arXiv:2501.06192}
}

read the original abstract

In the fields of computation and neuroscience, much is still unknown about the underlying computations that enable key cognitive functions including learning, memory, abstraction and behavior. This paper proposes a mathematical and computational model of learning and memory based on a small set of bio-plausible functions that include coincidence detection, signal modulation, and reward/penalty mechanisms. Our theoretical approach proposes that these basic functions are sufficient to establish and modulate an information space over which computation can be carried out, generating signal gradients usable for inference and behavior. The computational method used to test this is a structurally dynamic cellular automaton with continuous-valued cell states and a series of recursive steps propagating over an undirected graph with the memory function embedded entirely in the creation and modulation of graph edges. The experimental results show: that the toy model can make near-optimal choices to re-discover a reward state after a single training run; that it can avoid complex penalty configurations; that signal modulation and network plasticity can generate exploratory behaviors in sparse reward environments; that the model generates context-dependent memory representations; and that it exhibits high computational efficiency because of its minimal, single-pass training requirements combined with flexible and contextual memory representation.

Figures

Figures reproduced from arXiv: 2501.06192 by the authors.

**Figure 1.** Environment graph or external topology. 4.2 Reward and Penalty Navigation in Lattice Grids Testing the model in an 8x8, 64-node lattice graph approximating a reinforcement learning gridworld task showed that the agent could successfully navigate to the reward-connected node after a single training run through the environment graph (Fig 4a). Note that in this series of experiments, a reward node 65 is connected to th… view at source ↗

**Figure 2.** Coincident graph, or internal representation, after single training run. [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

**Figure 3.** Node values exhibit a gradient across a simple graph. In Fig.3 an input at location 13 generates a [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Figures from the paper (8 more)

**Figure 4.** Figure 4: Agent behavior with reward node 65 connected to node 64 (labeled R) in Fig.4a and two configura [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]

**Figure 5.** Figure 5: Navigating a more complex penalty environment. Fig 5a shows the path with no penalty nodes. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]

**Figure 6.** Figure 6: A simple demonstration of context-based (input vector) task-switching. Two rewards R1 and R2 at [PITH_FULL_IMAGE:figures/full_fig_p009_6.png]

**Figure 7.** Figure 7: Heatmap of node values at each step as the agent moves from start location node 1 to target node 64 [PITH_FULL_IMAGE:figures/full_fig_p009_7.png]

**Figure 8.** Figure 8: 100-node lattice graph, 400 random start/end points, showing histogram and box plot of path [PITH_FULL_IMAGE:figures/full_fig_p010_8.png]

**Figure 9.** Figure 9: 100-node, 0% randomized Watts-Strogatz graph, 400 random start/end points, showing histogram [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]

**Figure 10.** Figure 10: 100-node, 50% randomized Watts-Strogatz graph, 400 random start/end points, showing histogram [PITH_FULL_IMAGE:figures/full_fig_p011_10.png]

**Figure 11.** Figure 11: 100-node, 100 % randomized Watts-Strogatz graph, 400 random start/end points, showing success [PITH_FULL_IMAGE:figures/full_fig_p011_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 30 canonical work pages

[1]

E., Froemke, R

Aljadeff, J., D'amour, J., Field, R. E., Froemke, R. C., and Clopath, C. (2019). Cortical credit assignment by hebbian, neuromodulatory and inhibitory plasticity. arXiv preprint arXiv:1911.00307

arXiv 2019
[2]

Alonso-Sanz, R. (2007). A structurally dynamic cellular automaton with memory. Chaos, Solitons & Fractals , 32(4):1285--1295

work page 2007
[3]

and D Reyes, A

Barral, J. and D Reyes, A. (2016). Synaptic scaling rule preserves excitatory--inhibitory balance and salient neuronal network dynamics. Nature neuroscience , 19(12):1690--1696

work page 2016
[4]

W., Hamrick, J

Battaglia, P. W., Hamrick, J. B., Bapst, V., Sanchez-Gonzalez, A., Zambaldi, V., Malinowski, M., Tacchetti, A., Raposo, D., Santoro, A., Faulkner, R., et al. (2018). Relational inductive biases, deep learning, and graph networks. arXiv preprint arXiv:1806.01261

arXiv 2018
[5]

S., Matsumoto, M., and Hikosaka, O

Bromberg-Martin, E. S., Matsumoto, M., and Hikosaka, O. (2010). Dopamine in motivational control: rewarding, aversive, and alerting. Neuron , 68(5):815--834

work page 2010
[6]

Brown, T. H. and Chattarji, S. (1994). Hebbian synaptic plasticity: evolution of the contemporary concept. In Models of Neural Networks: Temporal Aspects of Coding and Information Processing in Biological Systems , pages 287--314. Springer New York New York, NY

work page 1994
[7]

Garner, A. R. and Keller, G. B. (2022). A cortical circuit for audio-visual predictions. Nature neuroscience , 25(1):98--105

work page 2022
[8]

S., Tanner, J., and Itti, L

Grant, W. S., Tanner, J., and Itti, L. (2017). Biologically plausible learning in neural networks with modulatory feedback. Neural Networks , 88:32--48

work page 2017

Show all 36 references

[9]

Grattarola, D., Livi, L., and Alippi, C. (2021). Learning graph cellular automata. Advances in Neural Information Processing Systems , 34:20983--20994

2021
[10]

Hassabis, D., Kumaran, D., Summerfield, C., and Botvinick, M. (2017). Neuroscience-inspired artificial intelligence. Neuron , 95(2):245--258

2017
[11]

Hebb, D. O. (2005). The organization of behavior: A neuropsychological theory

2005
[12]

Hopfield, J. J. (1982). Neural networks and physical systems with emergent collective computational abilities. Proceedings of the national academy of sciences , 79(8):2554--2558

1982
[13]

and Halpern, P

Ilachinski, A. and Halpern, P. (2009). Structurally dynamic cellular automata

2009
[14]

F., Vargas-Gutierrez, P., Okray, Z., Vietti-Michelina, S., Felsenberg, J., and Waddell, S

Jacob, P. F., Vargas-Gutierrez, P., Okray, Z., Vietti-Michelina, S., Felsenberg, J., and Waddell, S. (2021). Prior experience conditionally inhibits the expression of new learning in drosophila. Current biology , 31(16):3490--3503

2021
[15]

Kaneko, K. (1991). Coupled map lattice. In Chaos, Order, and Patterns , pages 237--247. Springer US Boston, MA

1991
[16]

P., Chakravarthi Raja, S., Cheney, N., Clune, J., et al

Kudithipudi, D., Aguilar-Simon, M., Babb, J., Bazhenov, M., Blackiston, D., Bongard, J., Brna, A. P., Chakravarthi Raja, S., Cheney, N., Clune, J., et al. (2022). Biological underpinnings for lifelong learning machines. Nature Machine Intelligence , 4(3):196--210

2022
[17]

J., Grimwood, P

Martin, S. J., Grimwood, P. D., and Morris, R. G. (2000). Synaptic plasticity and memory: an evaluation of the hypothesis. Annual review of neuroscience , 23(1):649--711

2000
[18]

Montague, P. R. and Sejnowski, T. J. (1994). The predictive brain: temporal coincidence and temporal order in synaptic learning mechanisms. Learning & Memory , 1(1):1--33

1994
[19]

F., Bridgeford, E

Muldoon, S. F., Bridgeford, E. W., and Bassett, D. S. (2016). Small-world propensity and weighted brain networks. Scientific reports , 6(1):22057

2016
[20]

Neumann, J. v. (1966). Theory of self-reproducing automata. Edited by Arthur W. Burks

1966
[21]

S., Park, Y.-G., Kim, M

Roy, D. S., Park, Y.-G., Kim, M. E., Zhang, Y., Ogawa, S. K., DiNapoli, N., Gu, X., Cho, J. H., Choi, H., Kamentsky, L., et al. (2022). Brain-wide mapping reveals that engrams for a single memory are distributed across multiple brain regions. Nature communications , 13(1):1799

2022
[22]

J., Roy, D

Ryan, T. J., Roy, D. S., Pignatelli, M., Arons, A., and Tonegawa, S. (2015). Engram cells retain memory under retrograde amnesia. Science , 348(6238):1007--1013

2015
[23]

P., Kim, H., and Chua, L

Sah, M. P., Kim, H., and Chua, L. O. (2014). Brains are made of memristors. IEEE circuits and systems magazine , 14(1):12--36

2014
[24]

Silver, D., Singh, S., Precup, D., and Sutton, R. S. (2021). Reward is enough. Artificial Intelligence , 299:103535

2021
[25]

L., and Barto, A

Singh, S., Lewis, R. L., and Barto, A. G. (2009). Where do rewards come from. In Proceedings of the annual conference of the cognitive science society , pages 2601--2606. Cognitive Science Society

2009
[26]

Sutton, R. S. and Barto, A. G. (2018). Reinforcement learning: An introduction . MIT press

2018
[27]

J., Jennings, K

Threlfell, S., Lalic, T., Platt, N. J., Jennings, K. A., Deisseroth, K., and Cragg, S. J. (2012). Striatal dopamine release is triggered by synchronized activity in cholinergic interneurons. Neuron , 75(1):58--64

2012
[28]

Todorov, E. (2009). Efficient computation of optimal actions. Proceedings of the national academy of sciences , 106(28):11478--11483

2009
[29]

and Margolus, N

Toffoli, T. and Margolus, N. (1987). Cellular automata machines: a new environment for modeling . MIT press

1987
[30]

S., and Ryan, T

Tonegawa, S., Pignatelli, M., Roy, D. S., and Ryan, T. J. (2015). Memory engram storage and retrieval. Current opinion in neurobiology , 35:101--109

2015
[31]

M., Siegelmann, H

Van de Ven, G. M., Siegelmann, H. T., and Tolias, A. S. (2020). Brain-inspired replay for continual learning with artificial neural networks. Nature communications , 11(1):4069

2020
[32]

Vardalaki, D., Chung, K., and Harnett, M. T. (2022). Filopodia are a structural substrate for silent synapses in adult neocortex. Nature , 612(7939):323--327

2022
[33]

Watts, D. J. and Strogatz, S. H. (1998). Collective dynamics of ‘small-world’networks. nature , 393(6684):440--442

1998
[34]

Williams, C. K. (2022). On suspicious coincidences and pointwise mutual information. Neural Computation , 34(10):2037--2046

2022
[35]

Wolfram, S. (1984). Cellular automata as models of complexity. Nature , 311(5985):419--424

1984
[36]

Zhang, T., Rosenberg, M., Jing, Z., Perona, P., and Meister, M. (2023). Endotaxis: A neuromorphic algorithm for mapping, goal-learning, navigation, and patrolling. eLife , 12

2023

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

E., Froemke, R

Aljadeff, J., D'amour, J., Field, R. E., Froemke, R. C., and Clopath, C. (2019). Cortical credit assignment by hebbian, neuromodulatory and inhibitory plasticity. arXiv preprint arXiv:1911.00307

arXiv 2019

[2] [2]

Alonso-Sanz, R. (2007). A structurally dynamic cellular automaton with memory. Chaos, Solitons & Fractals , 32(4):1285--1295

work page 2007

[3] [3]

and D Reyes, A

Barral, J. and D Reyes, A. (2016). Synaptic scaling rule preserves excitatory--inhibitory balance and salient neuronal network dynamics. Nature neuroscience , 19(12):1690--1696

work page 2016

[4] [4]

W., Hamrick, J

Battaglia, P. W., Hamrick, J. B., Bapst, V., Sanchez-Gonzalez, A., Zambaldi, V., Malinowski, M., Tacchetti, A., Raposo, D., Santoro, A., Faulkner, R., et al. (2018). Relational inductive biases, deep learning, and graph networks. arXiv preprint arXiv:1806.01261

arXiv 2018

[5] [5]

S., Matsumoto, M., and Hikosaka, O

Bromberg-Martin, E. S., Matsumoto, M., and Hikosaka, O. (2010). Dopamine in motivational control: rewarding, aversive, and alerting. Neuron , 68(5):815--834

work page 2010

[6] [6]

Brown, T. H. and Chattarji, S. (1994). Hebbian synaptic plasticity: evolution of the contemporary concept. In Models of Neural Networks: Temporal Aspects of Coding and Information Processing in Biological Systems , pages 287--314. Springer New York New York, NY

work page 1994

[7] [7]

Garner, A. R. and Keller, G. B. (2022). A cortical circuit for audio-visual predictions. Nature neuroscience , 25(1):98--105

work page 2022

[8] [8]

S., Tanner, J., and Itti, L

Grant, W. S., Tanner, J., and Itti, L. (2017). Biologically plausible learning in neural networks with modulatory feedback. Neural Networks , 88:32--48

work page 2017