Unveiling and Steering Connectome Organization with Interpretable Latent Variables

Guozhang Chen; Xingyu Liu; Yubin Li

REVIEW 5 major objections 8 minor 49 references

Unveiling and Steering Connectome Organization with Interpretable Latent Variables

T0 review · 5 major / 8 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 The paper argues that a compact set of latent variables can both encode and controllably steer the structure of fruit fly connectome subgraphs.

desk verdict A credible engineering pipeline for interpretable connectome latent spaces, but the 'predefined properties' claim only holds for edge count—still worth a serious referee. read the letter →

arxiv 2505.13011 v2 pith:PYODSI3F submitted 2025-05-19 cs.AI

Yubin Li , Xingyu Liu , Guozhang Chen This is my paper · ORCID

classification cs.AI

keywords connectomelatentrepresentationvariationalautoencoderSHAPgraphgenerationDrosophilaFlyWireinterpretability

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 tries to establish that the local wiring of the fruit fly brain is organized around a small number of low-dimensional principles, and that these principles can be captured by a variational autoencoder trained on subgraphs sampled from the FlyWire connectome. It claims that only a few latent dimensions carry most of the information about structural features such as edge count, reciprocity, betweenness centrality, and the fraction of non-neuronal cells, with SHAP analysis identifying exactly which dimensions matter. It further claims that by adjusting these latent codes, one can generate new connectome subgraphs with targeted structural properties, and it supports this with graph reconstruction results and controlled generation experiments. If correct, this gives neuroscience a tool for testing hypotheses about circuit design rules and offers a route toward bio-inspired artificial neural networks.

What carries the argument

The central machinery is the combination of an information-bottleneck variational autoencoder with a SHAP-based explainability module. The encoder uses multi-head graph attention networks to embed the directed adjacency matrix, concatenates these edge embeddings with one-hot node-type features, and maps them to a low-dimensional Gaussian latent $z$; the decoder reconstructs the adjacency matrix and node categories from $z$. A differentiable surrogate model approximates each graph statistic so that SHAP values can attribute the statistics to latent dimensions, and the additive property of SHAP values turns the task of hitting a target statistic into a knapsack-like dynamic programming problem. This pipeline is what carries the argument from "the latent space reconstructs the data" to "the latent space can be steered."

What would settle it

Train the same VAE and SHAP pipeline on 100-neuron subgraphs sampled randomly within the visual region without aligning to the visual-stream axis; if the set of influential latent dimensions and their mappings to edge count, reciprocity, betweenness, and non-neuronal ratio do not persist, the claimed organizational principles are an artifact of the cylindrical sampling. A second check is to compare subgraphs generated by the steering procedure against anatomically identified columnar units in the FlyWire dataset to see whether the controlled generations match real functional columns.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that connectome subgraphs from the fly visual system can be compressed into low-dimensional latent codes that preserve essential topology, and that individual latent dimensions are interpretably tied to measurable structural statistics. Using a VAE with an information bottleneck, the model reconstructs both the adjacency structure and the neurotransmitter-type labels of padded 100-node subgraphs, and SHAP analysis shows that a handful of dimensions (notably dimensions 1, 9, 11, 13, 24, and 28) dominate four structural readouts: edge count, reciprocity, betweenness, and non-neuronal ratio. The paper then demonstrates that manipulating these dimensions through SHAP-guided dynamic programming or CMA-ES produces graphs whose statistics move toward the requested targets, with an AUC of 0.919 when the full target graph is available and 0.852 when only its statistics are used.

Load-bearing premise

The load-bearing assumption is that the cylinder-sampled subgraphs are functionally meaningful columnar units of the fly visual system; if the sampling is arbitrary or biased, the learned latent variables and their apparent ties to structural features could reflect sampling artifacts rather than genuine brain organization.

Editorial extensions

If this is right

If the claim holds, local connectome structure can be analyzed and compared through compact latent codes rather than through full graph statistics, making cross-region and cross-species comparisons tractable.
The identified steering dimensions provide a direct way to test causal hypotheses: generate subgraphs with perturbed statistics and probe how those perturbations affect circuit function.
The same pipeline can be applied to other whole-brain connectomes as they are mapped, to test whether low-dimensional organizational principles generalize beyond the fly.
Controlled subgraph generation could supply synthetic neural-network training data with specified topology, bridging connectomics and artificial neural network design.
The reconstruction metrics quantify how much information survives the bottleneck, giving an empirical bound on the compressibility of local connectome structure.

Reading between the lines

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

My inference: the cylindrical sampling assumption could be tested directly by training the same VAE on randomly selected 100-neuron subgraphs that ignore the visual-stream axis; if the influential latent dimensions change, the discovered principles are sampling artifacts rather than columnar biology.
My inference: the genetic-bottleneck framing suggests a quantitative prediction the paper does not make: the information content of the six influential dimensions should scale sublinearly with subgraph neuron count, and measuring that scaling across subgraph sizes would test whether the bottleneck truly mirrors genome-scale compression.
My inference: because the SHAP-guided DP search treats dimensions additively, explicitly modeling pairwise interactions between the six influential dimensions could either improve steering accuracy or reveal that the paper's linear attribution masks coupled effects.
My inference: if the same six dimensions control the same statistics in other brain regions of the fly, that would support a universal "wiring code" rather than region-specific organization; this is a natural transfer experiment the paper does not run.

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: this is a credible engineering paper that assembles familiar pieces—GAT-VAE, SHAP, DP, CMA-ES—into a pipeline for Drosophila connectome subgraphs, and it reports its failures honestly. The headline claim, however, is broader than the evidence: the only target statistic that the SHAP-guided DP actually steers is edge count. Appendix H says so explicitly for reciprocity, betweenness, and non-neuronal. That matters more than the abstract suggests.

What's genuinely new is the unusual application of graph VAEs to spatially sampled connectome subgraphs with a SHAP-based interpretability overlay, and the demonstration that a small set of latent dimensions strongly controls edge count. Reconstruction is good (node acc 0.982, edge acc 0.978, edge AUC 0.959), and the RQ1 MMD table shows the model is competitive with diffusion baselines on degree and clustering. The surrogate models have high Pearson correlations, and the paper is admirably candid about the DP limitations in Appendix H.

Soft spots, in order of seriousness:

1. The DP procedure assumes SHAP values give a global additive decomposition, but SHAP's efficiency property holds only at a single input point. The paper presents no validation of the additivity assumption, and its own Appendix H shows the assumption breaks down.

2. No error bars or significance tests anywhere. MMD differences of 0.002 vs 0.004 may be noise; the table has no variance. This is fixable and necessary.

3. The sampling step—random XZ center, cylinder radius fit to get 81–100 neurons—gets no representativeness check. The latent structure may reflect the sampling rule rather than connectome organization.

4. Minor but real: '6 key dimensions' in one paragraph, '8 key dimensions' in the next; Appendix G's dimensionality analysis stops at 6, so why 8? Also, no code/data release, which hurts reproducibility.

The CMA-ES experiments are a reasonable workaround but lack baselines and error bars, so they don't rescue the uncontrolled claims.

Who is this for? Someone working in connectome representation learning or graph VAE interpretability will find useful ideas and an honest evaluation. It's not a foundational result. My take: it deserves a serious referee, not a desk reject, but only with major revisions—the abstract's controlled-generation claim needs to be scoped to edge count, or all four statistics need working DP, and the code should be released. I would not cite it for the 'genetic bottleneck' interpretation, but I might cite it as an application of interpretable GVAEs to connectomes.

Referee Report

5 major / 8 minor

Summary. This manuscript proposes an end-to-end framework, applied to the FlyWire Drosophila connectome, for learning interpretable low-dimensional representations of neural-circuit subgraphs and using those codes for controlled graph generation. The pipeline consists of (i) adaptive cylindrical sampling of 81-100-neuron subgraphs along the visual-stream axis, padded to 100 nodes; (ii) a graph VAE with a GAT-based edge encoder, a latent bottleneck, and a staged '1+2n' training protocol; (iii) a differentiable surrogate per graph statistic (edge count, reciprocity, betweenness, non-neuronal count) that enables SHAP attribution of latent dimensions; and (iv) two steering mechanisms, a SHAP-guided dynamic program (Algorithm 1) and CMA-ES. The experiments report competitive graph-generation MMD versus EDGE, DisCo, GDSS, and GruM; surrogate correlations between 0.92 and 0.998; a monotonic DP steering curve for edge count; and AUC of 0.919/0.852 in RQ3 when generating a target graph from full versus statistical information. Appendix H concedes that DP 'yields suboptimal results' for the other three statistics, attributing this to coupling between dimensions.

Significance. If the results hold, the framework is a novel and potentially useful combination of connectomics subgraph sampling, information-bottleneck representation learning, and SHAP-based white-box steering, and several parts of the manuscript are genuine strengths: the surrogate is validated against the ground-truth statistics of decoded graphs rather than only against its own training target, the experiments use the real FlyWire dataset, and the authors candidly report where DP fails. That candor, however, reveals that the headline claim--'controllably generate connectome subgraphs with predefined properties'--is demonstrated for one of four statistics via DP, while the complementary CMA-ES experiments lack baselines and variance estimates. The contribution to connectomics is also conditional on the unvalidated assumption that the cylinder-sampled subgraphs correspond to functionally meaningful columnar units. With a narrowed central claim, added statistical support, and validation of the steering assumptions, this could be a solid contribution to interpretable graph representation learning; in its current form the abstract overstates the demonstrated capability.

major comments (5)

[Abstract; Sec. 4.3; Appendix H] The manuscript's central load-bearing claim, stated in the abstract and conclusion, is that latent codes can be manipulated to 'controllably generate connectome subgraphs with predefined properties.' Appendix H reports that for reciprocity, betweenness, and non-neuronal count, dynamic programming 'yields suboptimal results,' so the DP-based controlled generation is demonstrated only for edge count (Figure 3c). The CMA-ES experiments in Sec. 4.4 report AUC of 0.919 and 0.852, but with no baselines, no comparison against optimizing latent vectors in a non-interpretable space of equal dimension, and no per-seed variability; they therefore cannot carry the full weight of the 'predefined properties' claim. The claims should be narrowed accordingly, or the missing baselines and variance information should be supplied.
[Sec. 3.5, Algorithm 1; Appendix E] Algorithm 1 assumes a global additive decomposition of each target statistic in terms of per-dimension SHAP values, expressed by the transition dp[i][j + shap[i][k]] <- k. SHAP values are local attributions: the efficiency property in Eq. (24) holds at a single input point and does not license treating shap[i][k] as a coordinate-independent contribution of setting z_i = k. The paper never tests this additivity assumption, and Appendix H's own diagnosis--that coupling between dimensions induces excessive variance in SHAP values--is direct internal evidence that the assumption fails for three of four statistics. The authors should either establish conditions under which Algorithm 1 is valid, validate additivity for the statistics it is applied to, or replace the DP with a steering method that accounts for interactions.
[Sec. 4.1; Tables 1-2; Sec. 4.4] No reported metric carries an error bar, confidence interval, or significance test. Table 1 supports comparative statements ('matching' GruM on degree MMD, 'particularly strong performance in clustering coefficient'), yet each cell appears to be a single value, and the baseline procedure of 'selecting the best-performing configuration' from 10 random hyperparameter samples is not an acceptable substitute for a repeated-seed protocol with reported variability. The surrogate correlations in Table 2 motivate the entire SHAP analysis; for reciprocity (0.9192) and betweenness (0.9622) the unexplained variance is substantial enough to matter for the downstream steering failure, so confidence intervals should be reported here as well.
[Sec. 3.1; Appendix D, Eq. (21)] The biological claims of the paper depend on the assertion that cylindrical sampling 'captures columnar functional units characteristic of neural circuitry,' but this is not validated: there is no comparison with anatomically defined columnar units, no sensitivity analysis for the cylinder radius or the 81-100 neuron window, and the centers are drawn randomly in the XZ-plane. Relatedly, the 'non-neuronal' count used as a target statistic is, by the paper's own definition in Eq. (21), the number of padding nodes introduced to standardize variable-size subgraphs to 100 nodes; it is an artifact of the sampling protocol rather than a structural property of the circuit. This should be acknowledged when the paper claims that latent dimensions 'govern important structural properties' of neural circuits.
[Sec. 4.3] The RQ2 analysis is internally inconsistent about its principal result: the text identifies six influential dimensions (1, 9, 11, 13, 24, and 28) and two sentences later refers to 'these 8 key dimensions.' Moreover, the paper provides no quantitative criterion for declaring a dimension 'key' (no |SHAP| threshold, no permutation or ablation test), so the claim that a few dimensions govern structural properties is not precisely defined. The number, identity, and selection rule for key dimensions should be stated consistently and justified.

minor comments (8)

[References] The reference list contains exact-duplicate entries: [7] and [8] (Chen et al., 2023), [18] and [19] (Hou et al., 2024), [30] and [31] (Loshchilov and Hutter, 2016), and [50] and [51] (Xu et al., 2024); these should be merged.
[Appendix D, Eq. (18)] The reciprocity definition is internally inconsistent: the left-hand side sum_{i<j} A_ij A_ji / sum_{i!=j} A_ij (1 - A_ji) equals (number of mutual pairs)/(total edges - 2*(mutual pairs)), whereas the right-hand side tr(A^2)/(||A||_0 - tr(A^2)) equals 2*(mutual pairs)/(total edges - 2*(mutual pairs)); the numerator differs by a factor of two, and the 1000x scaling mentioned in the appendix is not defined in the main text.
[Appendix I] The text 'Figure 3 demonstrates our model's graph reconstruction capability' is a mis-reference: Figure 3 shows SHAP and DP results for edge count, while the reconstruction panels are Figure 9.
[Sec. 4.1; Algorithm 1] Several quantities needed to reproduce the experiments are missing: the latent dimensionality d used in the main runs (Appendix G sweeps 4-32), the binarization threshold kappa, the DP bounds l and r, the discretization of the target statistic y, and the distribution whose standard deviation sigma is used to set the search range in Algorithm 1; the phrase 'possible eigen value bounds (l <= lambda <= r)' does not match the latent variable notation used elsewhere.
[Appendix B, Eq. (8)] The opening equality of the VAE derivation, log p(x;theta) = integral of q(z|x) log p(x) dz, is incorrect as written (for normalized q the integral equals log p(x)), and the rest of the derivation would benefit from a clean restatement.
[Sec. 3.3] The training-loss notation is inconsistent: L1, Lrec_2, Lkl_2, and L3 = Ledge_rec_3 + Lnode_rec_3 + Lkl_3 are introduced without defining the three components of L3; this should be aligned across Sec. 3.3 and Appendix F.
[Author affiliations] The affiliation line contains 'Sichuang, China,' which appears to be a typo for 'Sichuan, China.'
[Appendix I] Appendix I itself notes that the reconstruction results 'would benefit from comparison with standard reconstruction baselines in future work'; because the decoder's output distribution feeds all downstream SHAP and steering analyses, this missing comparison should be acknowledged as a limitation in the main text as well.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity; the controlled-generation claim is overbroad relative to the evidence, but no derivation reduces to its own inputs.

full rationale

The paper's pipeline is a self-contained learning system: subgraphs are sampled from FlyWire and padded, a VAE is trained to reconstruct them, four surrogate regressors are fitted to ground-truth statistics of decoded graphs, SHAP values are computed on the surrogate, and DP or CMA-ES searches latent codes for target statistics. No equation defines a target quantity in terms of the claimed conclusion. The surrogate Pearson correlations in Table 2 are fit-quality diagnostics, not predictions, and the paper explicitly says ground-truth statistics rather than surrogate outputs are used for the final analysis. The SHAP-DP route is an optimization: SHAP values serve as knapsack coefficients, and the reported achieved statistic comes from the decoded graph, so the monotonic edge-count curve in Figure 3c is an empirical result rather than an identity. Appendix H even states that DP 'yields suboptimal results' for reciprocity, betweenness, and non-neuronal, which would be impossible if the result were tautological. The 'genetic bottleneck' is motivational and does not enter any derivation. The one self-citation, reference [6] in the introduction, is a general background citation for 'structure dictates function' and is not load-bearing; no uniqueness theorem or ansatz is imported from prior work by the authors. The genuine weakness is evidential, not circular: the abstract's promise to 'controllably generate connectome subgraphs with predefined properties' is broader than the demonstrated DP success on edge count plus a baseline-free CMA-ES experiment. That is a correctness and support concern, not a circularity finding.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central claim rests on domain assumptions about the biological relevance of the sampling and latent space, plus unspecified hyperparameters. No entirely new entities are introduced. The free parameters and domain assumptions are the main things the paper does not independently justify.

free parameters (3)

Latent dimensionality d
The size of the latent bottleneck is central to the 'genetic bottleneck' claim but is chosen as a hyperparameter. The text refers to dimensions up to 28 and Appendix G tests 4 to 32, implying the main model uses 32, but this value is never justified.
Edge binarization threshold kappa
Used in Section 3.2 to convert probabilistic adjacency outputs to discrete values (Apred[i,j] = I(A'[i,j] >= kappa)). The value is not reported.
DP discretization step for SHAP values
Algorithm 1 loops over integer k from -sigma to sigma, requiring a discretization of the latent value range. The step size and bounds are not specified.

assumptions (5)

domain assumption A VAE with a low-dimensional latent space trained on connectome subgraphs can capture the brain's organizational principles analogous to the genetic bottleneck.
This is the core motivation in Section 1, linking genome information capacity to a 32-dim latent, but no derivation connects the two.
domain assumption Cylindrical sampling along the visual stream axis yields functionally relevant columnar units representative of the connectome.
Section 3.1 introduces the sampling strategy and asserts functional relevance without external validation.
domain assumption SHAP values computed on a surrogate model reflect the true relationship between latent variables and graph statistics.
The surrogate is validated by Pearson correlation in Table 2, but SHAP is applied to the surrogate, not the true generative process; the paper assumes the attribution transfers.
domain assumption The FlyWire dataset is an accurate reconstruction of the Drosophila visual connectome.
Used as ground truth throughout; the paper relies on the public dataset's correctness without independent verification.
standard math Standard VAE and SHAP mathematical properties hold.
Appendices B and E derive ELBO and SHAP using standard probability and game theory; these are accepted background results.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Unveiling and Steering Connectome Organization with Interpretable Latent Variables." pith.science (2026). https://pith.science/paper/PYODSI3F

@misc{pith2026250513011,
  author       = {Pith},
  title        = {Pith review of: Unveiling and Steering Connectome Organization with Interpretable Latent Variables},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PYODSI3F}},
  note         = {Machine review of arXiv:2505.13011}
}

read the original abstract

The brain's intricate connectome, a blueprint for its function, presents immense complexity, yet it arises from a compact genetic code, hinting at underlying low-dimensional organizational principles. This work bridges connectomics and representation learning to uncover these principles. We propose a framework that combines subgraph extraction from the Drosophila connectome, FlyWire, with a generative model to derive interpretable low-dimensional representations of neural circuitry. Crucially, an explainability module links these latent dimensions to specific structural features, offering insights into their functional relevance. We validate our approach by demonstrating effective graph reconstruction and, significantly, the ability to manipulate these latent codes to controllably generate connectome subgraphs with predefined properties. This research offers a novel tool for understanding brain architecture and a potential avenue for designing bio-inspired artificial neural networks.

Figures

Figures reproduced from arXiv: 2505.13011 by the authors.

**Figure 2.** Generation results on FlyWire in Adjacency Matrix Format. (a) Original: Shows the [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

**Figure 3.** Analysis of Edge Count.SHAP value results are shown in (a), where the y-axis represents [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Figures from the paper (6 more)

**Figure 4.** Figure 4: Reconstruction Capability Versus N.Figure (a) demonstrates the variation of edge recon [PITH_FULL_IMAGE:figures/full_fig_p016_4.png]

**Figure 5.** Figure 5: Model Reconstruction Capacity under Latent Code Variation.Figure (a) illustrates the [PITH_FULL_IMAGE:figures/full_fig_p017_5.png]

**Figure 6.** Figure 6: SHAP Values. (a) Betweenness (b) Non-Neuronal (c) Reciprocity [PITH_FULL_IMAGE:figures/full_fig_p018_6.png]

**Figure 7.** Figure 7: Dimension-Feature Relationships. (a) Betweenness (b) Non-Neuronal (c) Reciprocity [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]

**Figure 8.** Figure 8: DP Results. between dimensions induces excessive variance in SHAP values, thereby destabilizing the dynamic programming optimization. I Reconstruction Performance The reconstruction results in [PITH_FULL_IMAGE:figures/full_fig_p018_8.png]

**Figure 9.** Figure 9: Graph reconstruction results: Original samples (top) and their reconstructed counterparts [PITH_FULL_IMAGE:figures/full_fig_p019_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

49 extracted references · 28 canonical work pages

[1]

Machine learning explainability in nasopharyngeal cancer survival using lime and shap

Rasheed Omobolaji Alabi, Mohammed Elmusrati, Ilmo Leivo, Alhadi Almangush, and Antti A Mäkitie. Machine learning explainability in nasopharyngeal cancer survival using lime and shap. Scientific Reports, 13(1):8984, 2023

work page 2023
[2]

Tutorial cma-es: evolution strategies and covariance matrix adaptation

Anne Auger and Nikolaus Hansen. Tutorial cma-es: evolution strategies and covariance matrix adaptation. In Proceedings of the 14th annual conference companion on Genetic and evolutionary computation, pages 827–848, 2012

work page 2012
[3]

Dynamic programming

Richard Bellman. Dynamic programming. science, 153(3731):34–37, 1966

1966
[4]

From the gene to behavior

Seymour Benzer. From the gene to behavior. Jama, 218(7):1015–1022, 1971

work page 1971
[5]

Eg-nas: neural architecture search with fast evolutionary exploration

Zicheng Cai, Lei Chen, Peng Liu, Tongtao Ling, and Yutao Lai. Eg-nas: neural architecture search with fast evolutionary exploration. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 11159–11167, 2024. 9

work page 2024
[6]

A data-based large-scale model for primary visual cortex enables brain-like robust and versatile visual processing.Science advances, 8(44):eabq7592, 2022

Guozhang Chen, Franz Scherr, and Wolfgang Maass. A data-based large-scale model for primary visual cortex enables brain-like robust and versatile visual processing.Science advances, 8(44):eabq7592, 2022

work page 2022
[8]

Efficient and degree-guided graph generation via discrete diffusion modeling

Xiaohui Chen, Jiaxing He, Xu Han, and Li-Ping Liu. Efficient and degree-guided graph generation via discrete diffusion modeling. arXiv preprint arXiv:2305.04111, 2023

arXiv 2023
[9]

Flywire: online community for whole-brain connectomics

Sven Dorkenwald, Claire E McKellar, Thomas Macrina, Nico Kemnitz, Kisuk Lee, Ran Lu, Jingpeng Wu, Sergiy Popovych, Eric Mitchell, Barak Nehoran, et al. Flywire: online community for whole-brain connectomics. Nature methods, 19(1):119–128, 2022

work page 2022

Show all 49 references

[10]

What is dynamic programming? Nature biotechnology, 22(7):909–910, 2004

Sean R Eddy. What is dynamic programming? Nature biotechnology, 22(7):909–910, 2004

2004
[11]

Information theory and the genetic code

A Figureau. Information theory and the genetic code. Origins of Life and Evolution of the Biosphere, 17(3):439–449, 1987

1987
[12]

Sg-snn: a self-organizing spiking neural network based on temporal information

Shouwei Gao, Ruixin Zhu, Yu Qin, Wenyu Tang, and Hao Zhou. Sg-snn: a self-organizing spiking neural network based on temporal information. Cognitive Neurodynamics, 19(1):14, 2025

2025
[13]

Complex spiking neural networks with synaptic time-delay based on anti-interference function

Lei Guo, Sijia Zhang, Youxi Wu, and Guizhi Xu. Complex spiking neural networks with synaptic time-delay based on anti-interference function. Cognitive Neurodynamics, 16(6):1485–1503, 2022

2022
[14]

The cma evolution strategy: A tutorial

Nikolaus Hansen. The cma evolution strategy: A tutorial. arXiv preprint arXiv:1604.00772, 2016

2016 arXiv
[15]

Neuroscience-inspired artificial intelligence

Demis Hassabis, Dharshan Kumaran, Christopher Summerfield, and Matthew Botvinick. Neuroscience-inspired artificial intelligence. Neuron, 95(2):245–258, 2017

2017
[16]

Multi-level cognitive state classification of learners using complex brain networks and interpretable machine learning

Xiuling He, Yue Li, Xiong Xiao, Yingting Li, Jing Fang, and Ruijie Zhou. Multi-level cognitive state classification of learners using complex brain networks and interpretable machine learning. Cognitive Neurodynamics, 19(1):5, 2025

2025
[17]

beta-vae: Learning basic visual concepts with a constrained variational framework

Irina Higgins, Loic Matthey, Arka Pal, Christopher Burgess, Xavier Glorot, Matthew Botvinick, Shakir Mohamed, and Alexander Lerchner. beta-vae: Learning basic visual concepts with a constrained variational framework. In International conference on learning representations, 2017

2017
[19]

Improving molecular graph generation with flow matching and optimal transport

Xiaoyang Hou, Tian Zhu, Milong Ren, Dongbo Bu, Xin Gao, Chunming Zhang, and Shiwei Sun. Improving molecular graph generation with flow matching and optimal transport. arXiv preprint arXiv:2411.05676, 2024

2024 arXiv
[20]

Graph generation with diffusion mixture

Jaehyeong Jo, Dongki Kim, and Sung Ju Hwang. Graph generation with diffusion mixture. arXiv:2302.03596, 2024

2024 arXiv
[21]

Score-based generative modeling of graphs via the system of stochastic differential equations

Jaehyeong Jo, Seul Lee, and Sung Ju Hwang. Score-based generative modeling of graphs via the system of stochastic differential equations. arXiv:2202.02514, 2022

2022 arXiv
[22]

Generative ai: a review on models and applications

Kuldeep Singh Kaswan, Jagjit Singh Dhatterwal, Kiran Malik, and Anupam Baliyan. Generative ai: a review on models and applications. In 2023 International Conference on Communication, Security and Artificial Intelligence (ICCSAI), pages 699–704. IEEE, 2023

2023
[23]

Relevance factor vae: Learning and identifying disentangled factors

Minyoung Kim, Yuting Wang, Pritish Sahu, and Vladimir Pavlovic. Relevance factor vae: Learning and identifying disentangled factors. arXiv preprint arXiv:1902.01568, 2019. 10

1902 arXiv
[24]

Seong Hwan Kim, Hayom Kim, and Jung Bin Kim. Differences in functional network between focal onset nonconvulsive status epilepticus and toxic metabolic encephalopathy: application to machine learning models for differential diagnosis. Cognitive Neurodynamics, 17(4):845–853, 2023

2023
[25]

Auto-encoding variational bayes, 2013

Diederik P Kingma, Max Welling, et al. Auto-encoding variational bayes, 2013

2013
[26]

Variational inference of disentangled latent concepts from unlabeled observations

Abhishek Kumar, Prasanna Sattigeri, and Avinash Balakrishnan. Variational inference of disentangled latent concepts from unlabeled observations. arXiv preprint arXiv:1711.00848, 2017

2017 arXiv
[27]

Deep learning

Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. Deep learning. nature, 521(7553):436–444, 2015

2015
[28]

Dc-nas: Divide-and- conquer neural architecture search for multi-modal classification

Xinyan Liang, Pinhan Fu, Qian Guo, Keyin Zheng, and Yuhua Qian. Dc-nas: Divide-and- conquer neural architecture search for multi-modal classification. In Proceedings of the AAAI conference on artificial intelligence, volume 38, pages 13754–13762, 2024

2024
[29]

Challenging common assumptions in the unsupervised learning of disentangled representations

Francesco Locatello, Stefan Bauer, Mario Lucic, Gunnar Raetsch, Sylvain Gelly, Bernhard Schölkopf, and Olivier Bachem. Challenging common assumptions in the unsupervised learning of disentangled representations. In international conference on machine learning, pages 4114–
[31]

Cma-es for hyperparameter optimization of deep neural networks

Ilya Loshchilov and Frank Hutter. Cma-es for hyperparameter optimization of deep neural networks. arXiv preprint arXiv:1604.07269, 2016

2016 arXiv
[32]

A unified approach to interpreting model predictions

Scott M Lundberg and Su-In Lee. A unified approach to interpreting model predictions. Advances in neural information processing systems, 30, 2017

2017
[33]

Graphdf: A discrete flow model for molecular graph generation

Youzhi Luo, Keqiang Yan, and Shuiwang Ji. Graphdf: A discrete flow model for molecular graph generation. In International conference on machine learning, pages 7192–7203. PMLR, 2021

2021
[34]

Generative modelling of structurally constrained graphs.Advances in Neural Information Processing Systems, 37:137218– 137262, 2024

Manuel Madeira, Clement Vignac, Dorina Thanou, and Pascal Frossard. Generative modelling of structurally constrained graphs.Advances in Neural Information Processing Systems, 37:137218– 137262, 2024

2024
[35]

Toward an integration of deep learning and neuroscience

Adam H Marblestone, Greg Wayne, and Konrad P Kording. Toward an integration of deep learning and neuroscience. Frontiers in computational neuroscience, 10:94, 2016

2016
[36]

From explanations to feature selection: assessing shap values as feature selection mechanism

Wilson E Marcílio and Danilo M Eler. From explanations to feature selection: assessing shap values as feature selection mechanism. In 2020 33rd SIBGRAPI conference on Graphics, Patterns and Images (SIBGRAPI), pages 340–347. Ieee, 2020

2020
[37]

Algorithms for knapsack problems

Silvano Martello and Paolo Toth. Algorithms for knapsack problems. North-Holland Mathe- matics Studies, 132:213–257, 1987

1987
[38]

Overlooked implications of the reconstruction loss for vae disentanglement

Nathan Michlo, Richard Klein, and Steven James. Overlooked implications of the reconstruction loss for vae disentanglement. arXiv preprint arXiv:2202.13341, 2022

2022 arXiv
[39]

Computational role of structure in neural activity and connec- tivity

Srdjan Ostojic and Stefano Fusi. Computational role of structure in neural activity and connec- tivity. Trends in Cognitive Sciences, 2024

2024
[40]

Practical guide to shap analysis: explaining supervised machine learning model predictions in drug development

Ana Victoria Ponce-Bobadilla, Vanessa Schmitt, Corinna S Maier, Sven Mensing, and Sven Stodtmann. Practical guide to shap analysis: explaining supervised machine learning model predictions in drug development. Clinical and Translational Science, 17(11):e70056, 2024

2024
[41]

The importance of interpreting machine learning models for blood glucose prediction in diabetes: an analysis using shap

Francesco Prendin, Jacopo Pavan, Giacomo Cappon, Simone Del Favero, Giovanni Sparacino, and Andrea Facchinetti. The importance of interpreting machine learning models for blood glucose prediction in diabetes: an analysis using shap. Scientific reports, 13(1):16865, 2023. 11

2023
[42]

Detection of the chronic kidney disease using xgboost classifier and explaining the influence of the attributes on the model using shap

Md Johir Raihan, Md Al-Masrur Khan, Seong-Hoon Kee, and Abdullah-Al Nahid. Detection of the chronic kidney disease using xgboost classifier and explaining the influence of the attributes on the model using shap. Scientific Reports, 13(1):6263, 2023

2023
[43]

Connectome: How the brain’s wiring makes us who we are

Sebastian Seung. Connectome: How the brain’s wiring makes us who we are. HMH, 2012

2012
[44]

Graphaf: a flow-based autoregressive model for molecular graph generation

Chence Shi, Minkai Xu, Zhaocheng Zhu, Weinan Zhang, Ming Zhang, and Jian Tang. Graphaf: a flow-based autoregressive model for molecular graph generation. arXiv preprint arXiv:2001.09382, 2020

2001 arXiv
[45]

Encoding innate ability through a genomic bottleneck

Sergey Shuvaev, Divyansha Lachi, Alexei Koulakov, and Anthony Zador. Encoding innate ability through a genomic bottleneck. Proceedings of the National Academy of Sciences , 121(38):e2409160121, 2024

2024
[46]

Graphvae: Towards generation of small graphs using variational autoencoders

Martin Simonovsky and Nikos Komodakis. Graphvae: Towards generation of small graphs using variational autoencoders. In Artificial Neural Networks and Machine Learning–ICANN 2018: 27th International Conference on Artificial Neural Networks, Rhodes, Greece, October 4-7, 2018, Pr...

2018
[47]

Networks of the brain

Olaf Sporns, MARK DALEY , and JODY C CULHAM. Networks of the brain. Canadian Psychology, 52(4):321–322, 2011

2011
[48]

Bisnn: bio-information-fused spiking neural networks for enhanced eeg-based emotion recognition

Hongze Sun, Shifeng Mao, Wuque Cai, Yan Cui, Duo Chen, Dezhong Yao, and Daqing Guo. Bisnn: bio-information-fused spiking neural networks for enhanced eeg-based emotion recognition. Cognitive Neurodynamics, 19(1):1–12, 2025

2025
[49]

Graph attention networks

Petar Veliˇckovi´c, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Lio, and Yoshua Bengio. Graph attention networks. arXiv preprint arXiv:1710.10903, 2017

2017 arXiv
[51]

Discrete-state continuous-time diffusion for graph generation

Zhe Xu, Ruizhong Qiu, Yuzhong Chen, Huiyuan Chen, Xiran Fan, Menghai Pan, Zhichen Zeng, Mahashweta Das, and Hanghang Tong. Discrete-state continuous-time diffusion for graph generation. arXiv preprint arXiv:2405.11416, 2024

2024 arXiv
[52]

A critique of pure learning and what artificial neural networks can learn from animal brains

Anthony M Zador. A critique of pure learning and what artificial neural networks can learn from animal brains. Nature communications, 10(1):3770, 2019

2019
[53]

Pard: Permutation-invariant autoregressive diffusion for graph generation

Lingxiao Zhao, Xueying Ding, and Leman Akoglu. Pard: Permutation-invariant autoregressive diffusion for graph generation. arXiv preprint arXiv:2402.03687, 2024. A Graph Attention Network The Graph Attention Network (GAT) employs an attention mechanism to dynamically capture st...

2024 arXiv

Pith tools

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

[1] [1]

Machine learning explainability in nasopharyngeal cancer survival using lime and shap

Rasheed Omobolaji Alabi, Mohammed Elmusrati, Ilmo Leivo, Alhadi Almangush, and Antti A Mäkitie. Machine learning explainability in nasopharyngeal cancer survival using lime and shap. Scientific Reports, 13(1):8984, 2023

work page 2023

[2] [2]

Tutorial cma-es: evolution strategies and covariance matrix adaptation

Anne Auger and Nikolaus Hansen. Tutorial cma-es: evolution strategies and covariance matrix adaptation. In Proceedings of the 14th annual conference companion on Genetic and evolutionary computation, pages 827–848, 2012

work page 2012

[3] [3]

Dynamic programming

Richard Bellman. Dynamic programming. science, 153(3731):34–37, 1966

1966

[4] [4]

From the gene to behavior

Seymour Benzer. From the gene to behavior. Jama, 218(7):1015–1022, 1971

work page 1971

[5] [5]

Eg-nas: neural architecture search with fast evolutionary exploration

Zicheng Cai, Lei Chen, Peng Liu, Tongtao Ling, and Yutao Lai. Eg-nas: neural architecture search with fast evolutionary exploration. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, pages 11159–11167, 2024. 9

work page 2024

[6] [6]

A data-based large-scale model for primary visual cortex enables brain-like robust and versatile visual processing.Science advances, 8(44):eabq7592, 2022

Guozhang Chen, Franz Scherr, and Wolfgang Maass. A data-based large-scale model for primary visual cortex enables brain-like robust and versatile visual processing.Science advances, 8(44):eabq7592, 2022

work page 2022

[7] [8]

Efficient and degree-guided graph generation via discrete diffusion modeling

Xiaohui Chen, Jiaxing He, Xu Han, and Li-Ping Liu. Efficient and degree-guided graph generation via discrete diffusion modeling. arXiv preprint arXiv:2305.04111, 2023

arXiv 2023

[8] [9]

Flywire: online community for whole-brain connectomics

Sven Dorkenwald, Claire E McKellar, Thomas Macrina, Nico Kemnitz, Kisuk Lee, Ran Lu, Jingpeng Wu, Sergiy Popovych, Eric Mitchell, Barak Nehoran, et al. Flywire: online community for whole-brain connectomics. Nature methods, 19(1):119–128, 2022

work page 2022