BrainMAP: Learning Multiple Activation Pathways in Brain Networks

Aiying Zhang; Chen Chen; Guorong Wu; Jiaqi Ding; Jundong Li; Song Wang; Tianlong Chen; Xinyu Zhao; Yushun Dong; Zhen Tan

REVIEW 4 major objections 6 minor 63 references

BrainMAP: Learning Multiple Activation Pathways in Brain Networks

T0 review · 4 major / 6 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 BrainMAP models brain graphs as ordered activation pathways and beats prior fMRI predictors.

desk verdict The architecture is a genuine novelty and the accuracy gains are believable, but the 'activation pathway' interpretation is not established—the learned orders are shaped by task loss, not validated against any neural pathway ground truth. read the letter →

arxiv 2412.17404 v2 pith:3SBULAH2 submitted 2024-12-23 cs.AI

Song Wang , Zhenyu Lei , Zhen Tan , Jiaqi Ding , Xinyu Zhao , Yushun Dong , Guorong Wu , Tianlong Chen

show 3 more authors

Chen Chen Aiying Zhang Jundong Li

This is my paper · ORCID

classification cs.AI

keywords fMRIfunctionalconnectivitygraphsactivationpathwaysgraphsequentializationmixtureofexpertslong-rangedependenciesbrainnetworkanalysisMamba

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

BrainMAP is a framework for predicting behavior and demographics from fMRI-derived functional-connectivity graphs. Instead of relying on message passing over static edges, it learns an ordering of brain regions that approximates sequential information flow, feeds the ordered sequence to a mixture of sequential models, and aggregates the results. The paper argues that this captures multiple long-range activation pathways—chains of brain regions that co-activate during a task—which conventional graph models miss. On five real-world fMRI benchmarks it outperforms all baselines, and its saliency maps rank known task-relevant brain regions higher than prior models. The payoff is both better prediction and a more interpretable account of which brain regions work together for a task.

What carries the argument

Adaptive Graph Sequentialization: an order-learning graph neural network produces per-node scores; sorting these scores yields a candidate activation order. Because real ranks break differentiability, the paper standardizes scores to share the first two moments of a true rank variable (Theorem 4.1) and trains the order with a loss that is equivalent to maximizing Spearman rank correlation with sampled low-loss orders and minimizing it with high-loss orders (Theorem 4.2). Hierarchical Pathway Integration: for each learned order, a top-k gated mixture of P expert sequence models (selective state-space models such as Mamba) captures diverse pathways; the gating function uses self-attention and positional encoding so similar pathways consistently route to the same expert, and a weighted sum over M orders produces the final graph embedding.

What would settle it

Take an fMRI dataset with independently measured activation order (for example, from BOLD latency or effective connectivity) for a set of tasks, train BrainMAP, and compare its learned orders with the gold-standard orders under Spearman correlation. If the learned orders are no more aligned than random orders of the same regions, the pathway interpretation is falsified. A second check: randomize the learned orders while approximately preserving task loss; if accuracy does not drop, the sequentialization itself is not carrying the claimed signal.

Watch

Extended reading notes

Core claim

The central discovery the paper tries to establish is that a functional-connectivity graph can be re-expressed as several node sequences whose order encodes the direction of neural signal flow, and that this re-expression makes long-range multi-pathway structure learnable. The order is not fixed: an order-learning graph neural network assigns each brain region a score, and a contrastive objective pulls the resulting order toward sampled good orders (those with low task loss) and away from bad orders. A hierarchical mixture of sequence models then extracts multiple pathways within each order and merges representations across orders. The paper reports that this design outperforms all baselines on five HCP prediction tasks, that the mixture-of-experts component is the largest single contributor in ablations, and that the model's saliency maps identify expert-defined motor-task regions better than comparison models.

Load-bearing premise

The load-bearing premise is that an order of brain regions which lowers the task-prediction loss is genuinely closer to the order in which neural signals travel; if task loss is not a faithful proxy for pathway structure, the activation pathway interpretation is unsupported even if prediction accuracy still improves.

Editorial extensions

If this is right

On five HCP fMRI prediction tasks, BrainMAP outperforms all baselines, with improvements up to 4.09% over the strongest prior model and gains of up to 12.13% over traditional graph neural networks on HCP-Task.
Ablation studies show that both the learned sequentialization and the mixture-of-experts aggregation contribute to performance, with the mixture-of-experts being the most critical component and the load-balancing loss also helping.
For the MOTOR task, BrainMAP's saliency regions achieve Hit@10 of 19.38 and MRR of 9.26, outperforming the strongest sequence baseline (Hit@10 of 15.00 and MRR of 6.27), indicating better alignment with expert-defined activated brain regions.
Performance improves as the number of experts increases up to three and roughly plateaus beyond four, suggesting a limited number of latent activation pathways in the brain.
BrainMAP's average training time per epoch is higher than all baselines, ranging from about 9.23 to 34.93 seconds across datasets on four A100 GPUs.

Reading between the lines

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

If the learned order truly approximates neural activation order, the same sequentialization-plus-mixture recipe could transfer to other domains where signal propagation follows ordered paths, such as gene regulatory cascades, epidemic spread, or circuit traces.
The order learner uses task loss as a proxy for pathway structure, so a stronger test would compare learned orders against gold-standard effective-connectivity or BOLD-latency orderings; if they diverge, the accuracy gains might come from a sequential inductive bias rather than faithful pathway discovery.
The interpretation results suggest BrainMAP could generate candidate brain regions for tasks where expert ground truth is unknown, and those candidates could then be validated with targeted fMRI experiments.
Because expert count saturates around four, the gating distribution could be used to estimate the effective number of pathways per task and check whether that count aligns with known parallel pathway organization, such as the dorsal and ventral visual streams.

Share X Bluesky LinkedIn Reddit HN

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

Good paper to know about, but don't take the 'activation pathway' language at face value. The architecture is a genuine novelty — adaptive graph sequentialization with learned permutations, fused with a mixture of Mamba experts — and the empirical work is mostly honest: five HCP datasets, ablations, and a correct derivation showing the contrastive loss matches Spearman rank correlation. The accuracy gains are modest but consistent, and the method beats Graph-Mamba on every reported dataset.

The soft spot is exactly where the stress-test points. The order learner is trained entirely by the model's own task loss: sample orders, keep the low-loss ones as positives and high-loss ones as negatives, and pull the learned order toward the positives. That is a way to find convenient permutations for the sequence model, but nothing in the training ties those permutations to neural activation pathways. No ground truth, no check that consecutive nodes in a learned order are connected by significant edges, no comparison with known anatomical pathways. So the central claim — that BrainMAP 'extracts long-range dependencies along multiple activation pathways' — is not established. What is established is that multiple learned orders, each processed by a gated expert, improve prediction. That's a useful engineering result, but it is not a neuroscience result.

Minor issues are easy to list: MRR values in Table 4 are greater than 1 and the metric is undefined; Table 3 swaps the HCP-FI and HCP-WM columns; the abstract in the arXiv version points to the Graph-Mamba repo instead of the BrainMAP repo; there are no significance tests, and on HCP-Age and HCP-Gender the standard deviations overlap. None of these are load-bearing if fixed, but together they make the reported numbers less trustworthy than they should be.

The theoretical part is solid: the standardization trick in Eq. 4 and the equivalence proof in Theorem 4.2 are correct and genuinely useful for anyone working on differentiable sorting. The ablation shows both the learned order and the MoE matter, which supports the architecture's internal logic.

For a reader: if you work on graph sequence models or connectomics, this is worth a slot in a reading group, mainly as an example of how to make permutation learning work with task-loss feedback. It should be cited for the method, not for the pathway claim. I would send it to peer review — a good reviewer will ask for a validation of the pathway interpretation, or a softening of the language, plus the mechanical fixes. The core idea is strong enough to deserve that round.

Referee Report

4 major / 6 minor

Summary. The paper proposes BrainMAP, a framework for prediction on fMRI functional-connectivity graphs. BrainMAP first learns one or more node orderings via an order-learning GNN, then feeds each ordering to a small set of top- gated sequential models (Mamba or Transformer experts) and aggregates the resulting representations. The authors claim that this procedure extracts 'multiple activation pathways' in brain networks, yields state-of-the-art accuracy on five HCP datasets, and provides saliency-based explanations that rank task-relevant brain regions. The paper includes ablations, an efficiency study, and proofs for two theorems about the approximate-rank contrastive loss.

Significance. If the empirical results are sound, the paper offers a practical recipe for combining adaptive graph sequentialization with mixture-of-experts sequence models, and it addresses a real limitation of message-passing GNNs on long-range brain-network dependencies. The paper is also commendable for providing code, evaluating on five datasets, running ablations, and attempting an interpretation study. However, the central conceptual claim that the method learns 'activation pathways' is not validated: the order learner is optimized purely through the predictor's task loss, and the explanation experiments only measure node-level saliency. The predictive gains are also not supported by significance testing, and the reported MRR values are impossible as written. The contribution is therefore best viewed as an accuracy-oriented sequentialization technique whose pathway interpretation needs either direct validation or substantial softening.

major comments (4)

[§4.1, Eqs. (3) and (6); §5.4] The paper's central claim that BrainMAP learns 'activation pathways' is not supported by any validation of the learned orders against known neural pathways. The order learner is trained by contrasting sampled permutations that happen to give low downstream task loss (the 'positive' set) with permutations that give high task loss (the 'negative' set); the only supervision signal is the predictor's own loss, not any pathway ground truth. Moreover, each expert consumes a full linear ordering of all N nodes, so 'multiple pathways' are never enumerated or verified as graph paths. The explanation study in §5.4 only aggregates node saliency scores; it never compares the learned order to known anatomical or functional pathways. At minimum, the pathway-language claims should be softened, or the authors should provide a direct validation that learned orders correspond to plausible activation pathways (e.g., agreement with known task-evoked networks or effective-connectivity estimates).
[§5.2, Table 2] The headline 'outperforms all baselines' claim is not statistically supported. All experiments used only three seeds and no significance tests are reported, and under the reported standard deviations the BrainMAP mean overlaps with the best baseline on every dataset (e.g., HCP-Age: 48.44±1.65 vs. 46.35±2.73; HCP-Gender: 78.92±0.49 vs. 77.16±3.13; HCP-WM: 3.81±0.03 vs. 3.94±0.14). The authors should report paired significance tests across seeds or across subjects, or otherwise demonstrate that the improvements are not within noise.
[§5.4, Table 4] The MRR values in Table 4 (3.07, 3.13, 6.27, 9.26) are impossible for a Mean Reciprocal Rank, which is bounded above by 1. Either the metric is misnamed or the computation is incorrect, and since this table is the quantitative evidence for the explanation claim, the results need to be recomputed and the metric defined precisely. The text also does not specify which 'Mamba-specific explanation method' was used or how the ground-truth activated regions were obtained.
[§5.1, Appendix C] The experimental protocol does not describe how data are split into train, validation, and test sets. Appendix C states that hyperparameters are chosen by grid search and each configuration is run for 100 epochs, but it does not state whether the reported test numbers correspond to the best validation configuration or to the best test configuration. Without a clear split and selection protocol, the reported superiority could be optimistic. Please specify the split procedure, number of folds or seeds, and how early stopping or model selection was performed.

minor comments (6)

[§5.2] The text refers to 'Table 5' for the main results, but the main results are in Table 2; the reference should be corrected.
[Table 3] The rows for HCP-FI and HCP-WM appear to be switched relative to Table 2: Table 2 reports HCP-WM as 3.81 and HCP-FI as 10.75, while Table 3 reports HCP-FI as 3.81 and HCP-WM as 10.75. The column labels and values should be checked.
[Appendix B, after Eq. (26)] The proof of Theorem 4.2 states that Spearman's r satisfies 0 ≤ rs ≤ 1; this is false, since Spearman's rank correlation can be negative. The subsequent argument only needs rs < 1, so the proof can be repaired, but the claim should be corrected.
[§4.1, Eqs. (3) and (6)] The contrastive objective in Eq. (3) includes a weight λ that does not appear in the actual loss in Eq. (6). The relationship between these two formulations should be explained.
[§5.3, Table 3] The abbreviations 'w/o LR' and 'w/o LB' are used without definition; the text should state that LR refers to the order learner and LB to the load-balancing loss.
[Abstract and body] The abstract gives the code URL as https://github.com/LzyFischer/Graph-Mamba, while the body gives https://github.com/LzyFischer/BrainMAP; the correct link should be used consistently.

Circularity Check

1 steps flagged · score 3.0 of 10

One circular validation step: MoE expert diversity is renamed as evidence of 'multiple activation pathways'; the core benchmark prediction remains non-circular.

renaming known result [Section 5.5 (MoE Analysis), Fig. 5 discussion; cf. Section 4.2, Eqs. 9-12]
"The results illustrated that BrainMAP consistently maintains high activation rates, with a minimum of 66.7% activation rate for HCP-Gender. The findings further suggest that BrainMAP effectively extracts multiple pathways, as evidenced by the activation of diverse experts."

The paper's 'multiple activation pathways' are never defined independently of the architecture; in Section 4.2, each expert is described as capturing a specific type of pathway and the gating function (Eqs. 10-11) is designed to route 'similar pathways' to the same expert. The evidence offered in Section 5.5 is that different experts are activated across layers. But diverse activation of experts is a direct, by-construction consequence of the top-K gating mechanism and the load-balancing loss, not an external observation about neural pathways. Therefore the conclusion 'BrainMAP effectively extracts multiple pathways' reduces to the fact that multiple experts are active—i.e., the model's own mechanism is renamed as the neuroscientific construct it was designed to represent.

full rationale

The main predictive pipeline—adaptive graph sequentialization plus MoE with Mamba experts—is trained end-to-end on task labels and evaluated on held-out HCP data against external baselines, so the core performance claim (Section 5.2, Table 2) does not reduce to a fit or to a self-citation. The order-learning GNN does use the model's own task loss to select positive and negative permutations (Eqs. 3 and 6), but this is a legitimate reward signal for learning a permutation policy; the final predictor must still generalize and is measured on unseen labels. The mathematical theorems (4.1 and 4.2) are proven from definitions and do not smuggle in the target result. Self-citations to the authors' prior contrastive learning and MoE work are background, not load-bearing. The one genuinely circular element is the validation of the 'multiple activation pathways' interpretation in Section 5.5, where diverse expert activation—an architectural property of the MoE gating—is cited as evidence that multiple neural pathways are extracted. Since 'pathway' is operationalized only as an expert's sub-sequence, that evidence is equivalent to the construct by construction. This circularity affects the interpretive pathway claim, not the benchmark accuracy claim, so the overall circularity score is moderate.

Assumptions & free parameters 5 free parameters · 5 assumptions · 1 invented entities

The central architectural claims rest on several free hyperparameters and on the assumption that task-loss-minimizing node orders correspond to real activation pathways. The 'activation pathway' is an invented latent construct without independent falsifiable support.

free parameters (5)

M (number of orders) = 3
Set by grid search; a key architectural choice that affects performance (Appendix C).
K (number of top experts) = 2
Set by default; varied in Fig. 4 to analyze MoE behavior.
lambda (contrastive weight) = not specified
Used in Eq. 3 to balance positive and negative order similarities; not reported in the text.
Np/Nd (number of positive/negative sampled orders) = not specified
Sampling sizes for the order contrastive loss (Eq. 6); not specified in the paper.
learning rate and weight decay = grid searched from 1e-1 to 1e-3 and 1e-3 to 1e-5
Optimization hyperparameters selected by grid search (Appendix C), but no explicit validation split mentioned.

assumptions (5)

standard math Standard statistical identities for rank variables and Spearman correlation (Theorems 4.1 and 4.2).
Proofs in Appendices A and B rely on well-known properties of means, variances, and covariance of ranks.
domain assumption Human brain tasks involve activation pathways representable as paths over FC graphs.
Motivates the entire framework (Section 1, citing Sporns 2011).
domain assumption FC graphs contain sequential dependencies that can be captured by ordering nodes.
The adaptive sequentialization assumes that a permutation of nodes can preserve pathway information (Section 4.1).
ad hoc to paper The task loss is a valid proxy for the quality of a node order.
Used to select positive and negative orders in Eq. 3; if false, order learning would not promote meaningful pathways.
ad hoc to paper Top-k MoE gating can separate distinct activation pathways.
The gating function (Eq. 10) is assumed to specialize experts by pathway type, but this is not directly validated.

invented entities (1)

Activation pathway (latent sequential brain-region structure)
purpose: Motivates the architecture and provides a post-hoc interpretation of predictions.
The paper provides no direct measurement or ground-truth pathway labels. The explanation study only compares saliency of individual ROIs against expert-defined active regions, not pathway-level structure.

how reviews work

0 comments

Cite this review

Pith. "Pith review of BrainMAP: Learning Multiple Activation Pathways in Brain Networks." pith.science (2026). https://pith.science/paper/3SBULAH2

@misc{pith2026241217404,
  author       = {Pith},
  title        = {Pith review of: BrainMAP: Learning Multiple Activation Pathways in Brain Networks},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3SBULAH2}},
  note         = {Machine review of arXiv:2412.17404}
}

read the original abstract

Functional Magnetic Resonance Image (fMRI) is commonly employed to study human brain activity, since it offers insight into the relationship between functional fluctuations and human behavior. To enhance analysis and comprehension of brain activity, Graph Neural Networks (GNNs) have been widely applied to the analysis of functional connectivities (FC) derived from fMRI data, due to their ability to capture the synergistic interactions among brain regions. However, in the human brain, performing complex tasks typically involves the activation of certain pathways, which could be represented as paths across graphs. As such, conventional GNNs struggle to learn from these pathways due to the long-range dependencies of multiple pathways. To address these challenges, we introduce a novel framework BrainMAP to learn Multiple Activation Pathways in Brain networks. BrainMAP leverages sequential models to identify long-range correlations among sequentialized brain regions and incorporates an aggregation module based on Mixture of Experts (MoE) to learn from multiple pathways. Our comprehensive experiments highlight BrainMAP's superior performance. Furthermore, our framework enables explanatory analyses of crucial brain regions involved in tasks. Our code is provided at https://github.com/LzyFischer/Graph-Mamba.

Figures

Figures reproduced from arXiv: 2412.17404 by the authors.

**Figure 2.** The overall process of BrainMAP. We first adaptively learn [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** Interpretation results of BrainMAP for the task [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗

Figures from the paper (1 more)

**Figure 5.** Figure 5: The activation distribution of the experts across [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

63 extracted references · 50 canonical work pages

[1]

, " * write output.state after.block = add.period write newline

ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...
[2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...
[3]

S.; and Sporns, O

Bassett, D. S.; and Sporns, O. 2017. Network neuroscience. Nature neuroscience, 20(3): 353--364

work page 2017
[4]

Chang, C.-H.; Lin, C.-H.; and Lane, H.-Y. 2021. Machine learning and novel biomarkers for the diagnosis of Alzheimer’s disease. International journal of molecular sciences, 22(5): 2761

work page 2021
[5]

Chen, J.; Qi, Y.; Wang, Y.; and Pan, G. 2024. Bridging the Semantic Latent Space between Brain and Machine: Similarity Is All You Need. AAAI, 38: 11302--11310

work page 2024
[6]

Cui, H.; Dai, W.; Zhu, Y.; Kan, X.; Gu, A. A. C.; Lukemire, J.; Zhan, L.; He, L.; Guo, Y.; and Yang, C. 2022 a . Braingb: a benchmark for brain network analysis with graph neural networks. IEEE transactions on medical imaging

work page 2022
[7]

Cui, H.; Lu, Z.; Li, P.; and Yang, C. 2022 b . On positional and structural node features for graph neural networks on non-attributed graphs. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management, 3898--3902

work page 2022
[8]

Z.; Rueckert, D.; and Robinson, E

Dahan, S.; Williams, L. Z.; Rueckert, D.; and Robinson, E. C. 2021. Improving phenotype prediction using long-range spatio-temporal dynamics of functional connectivity. In Machine Learning in Clinical Neuroimaging: 4th International Workshop

work page 2021

Show all 63 references

[9]

N.; Fan, A.; Auli, M.; and Grangier, D

Dauphin, Y. N.; Fan, A.; Auli, M.; and Grangier, D. 2017. Language modeling with gated convolutional networks. In International conference on machine learning

2017
[10]

D.; Aghaeepour, N.; Ahn, A

Davis, K. D.; Aghaeepour, N.; Ahn, A. H.; Angst, M. S.; Borsook, D.; Brenton, A.; Burczynski, M. E.; Crean, C.; Edwards, R.; Gaudilliere, B.; et al. 2020. Discovery and validation of biomarkers to aid the development of safe and effective pain therapeutics: challenges and oppo...

2020
[11]

R.; and Saeed, F

Eslami, T.; Mirjalili, V.; Fong, A.; Laird, A. R.; and Saeed, F. 2019. ASD-DiagNet: a hybrid learning approach for detection of autism spectrum disorder using fMRI data. Frontiers in neuroinformatics, 13: 70

2019
[12]

S.; Poldrack, R

Finn, E. S.; Poldrack, R. A.; and Shine, J. M. 2023. Functional neuroimaging as a catalyst for integrated neuroscience. Nature, 623(7986): 263--273

2023
[13]

D.; and Raichle, M

Fox, M. D.; and Raichle, M. E. 2007. Spontaneous fluctuations in brain activity observed with functional magnetic resonance imaging. Nature reviews neuroscience

2007
[14]

Gao, D.; Zhang, H.; Li, P.; Tang, T.; Liu, S.; Zhou, Z.; Ying, S.; Zhu, Y.; and Zhang, Y. 2024. A Local-Ascending-Global Learning Strategy for Brain-Computer Interface. In AAAI, volume 38, 10039--10047

2024
[15]

Gu, A.; and Dao, T. 2023. Mamba: Linear-time sequence modeling with selective state spaces. arXiv preprint arXiv:2312.00752

2023 arXiv
[16]

J.; Nguyen, M.; Sabuncu, M

He, T.; Kong, R.; Holmes, A. J.; Nguyen, M.; Sabuncu, M. R.; Eickhoff, S. B.; Bzdok, D.; Feng, J.; and Yeo, B. T. 2020. Deep neural networks and kernel regression achieve comparable accuracies for functional connectivity prediction of behavior and demographics. NeuroImage

2020
[17]

Hsu, C.-Y.; Cox, K.; Xu, J.; Tan, Z.; Zhai, T.; Hu, M.; Pratt, D.; Chen, T.; Hu, Z.; and Ding, Y. 2024. Thought Graph: Generating Thought Process for Biological Reasoning. In Companion Proceedings of the ACM on Web Conference 2024, 537--540

2024
[18]

A.; Jordan, M

Jacobs, R. A.; Jordan, M. I.; Nowlan, S. J.; and Hinton, G. E. 1991. Adaptive mixtures of local experts. Neural computation, 3(1): 79--87

1991
[19]

Q.; Sablayrolles, A.; Roux, A.; Mensch, A.; Savary, B.; Bamford, C.; Chaplot, D

Jiang, A. Q.; Sablayrolles, A.; Roux, A.; Mensch, A.; Savary, B.; Bamford, C.; Chaplot, D. S.; Casas, D. d. l.; Hanna, E. B.; Bressand, F.; et al. 2024. Mixtral of experts. arXiv preprint arXiv:2401.04088

2024 arXiv
[20]

Jo, T.; Nho, K.; and Saykin, A. J. 2019. Deep learning in Alzheimer's disease: diagnostic classification and prognostic prediction using neuroimaging data. Frontiers in aging neuroscience, 11: 220

2019
[21]

I.; and Jacobs, R

Jordan, M. I.; and Jacobs, R. A. 1994. Hierarchical mixtures of experts and the EM algorithm. Neural computation

1994
[22]

Kan, X.; Dai, W.; Cui, H.; Zhang, Z.; Guo, Y.; and Yang, C. 2022. Brain network transformer. NeurIPS, 35: 25586--25599

2022
[23]

J.; Miller, S

Kawahara, J.; Brown, C. J.; Miller, S. P.; Booth, B. G.; Chau, V.; Grunau, R. E.; Zwicker, J. G.; and Hamarneh, G. 2017. BrainNetCNN: Convolutional neural networks for brain networks; towards predicting neurodevelopment. NeuroImage, 146: 1038--1049

2017
[24]

Keller, M.; Taube, W.; and Lauber, B. 2018. Task-dependent activation of distinct fast and slow (er) motor pathways during motor imagery. Brain stimulation, 11(4): 782--788

2018
[25]

C.; and Kim, J.-J

Kim, B.-H.; Ye, J. C.; and Kim, J.-J. 2021. Learning dynamic graph representation of brain connectome with spatio-temporal attention. NeurIPS, 34: 4314--4327

2021
[26]

N.; and Welling, M

Kipf, T. N.; and Welling, M. 2017. Semi-supervised classification with graph convolutional networks. In ICLR

2017
[27]

D.; and Woo, C.-W

Kohoutov \'a , L.; Heo, J.; Cha, S.; Lee, S.; Moon, T.; Wager, T. D.; and Woo, C.-W. 2020. Toward a unified framework for interpreting machine-learning models in neuroimaging. Nature protocols, 15(4): 1399--1435

2020
[28]

A.; Bonin, V.; Reed, M.; Graham, B

Lee, W.-C. A.; Bonin, V.; Reed, M.; Graham, B. J.; Hood, G.; Glattfelder, K.; and Reid, R. C. 2016. Anatomy and function of an excitatory network in the visual cortex. Nature, 532(7599): 370--374

2016
[29]

Li, H.; and Fan, Y. 2019. Interpretable, highly accurate brain decoding of subtly distinct brain states from functional MRI using intrinsic functional networks and long short-term memory recurrent neural networks. NeuroImage

2019
[30]

H.; Ventola, P.; and Duncan, J

Li, X.; Zhou, Y.; Dvornek, N.; Zhang, M.; Gao, S.; Zhuang, J.; Scheinost, D.; Staib, L. H.; Ventola, P.; and Duncan, J. S. 2021. Braingnn: Interpretable brain graph neural network for fmri analysis. Medical Image Analysis, 74: 102233

2021
[31]

Li, Y.; Zhang, X.; Nie, J.; Zhang, G.; Fang, R.; Xu, X.; Wu, Z.; Hu, D.; Wang, L.; Zhang, H.; et al. 2022. Brain connectivity based graph convolutional networks and its application to infant age prediction. IEEE transactions on medical imaging, 41(10): 2764--2776

2022
[32]

Liao, M.; Wan, G.; and Du, B. 2024. Joint Learning Neuronal Skeleton and Brain Circuit Topology with Permutation Invariant Encoders for Neuron Classification. In AAAI

2024
[33]

Liu, X.; Zhou, M.; Shi, G.; Du, Y.; Zhao, L.; Wu, Z.; Liu, D.; Liu, T.; and Hu, X. 2023. Coupling artificial neurons in bert and biological neurons in the human brain. In AAAI

2023
[34]

Luo, D.; Cheng, W.; Xu, D.; Yu, W.; Zong, B.; Chen, H.; and Zhang, X. 2020. Parameterized explainer for graph neural network. NeurIPS, 33: 19620--19631

2020
[35]

L.; Lenssen, J

Morris, C.; Ritzert, M.; Fey, M.; Hamilton, W. L.; Lenssen, J. E.; Rattan, G.; and Grohe, M. 2019. Weisfeiler and leman go neural: Higher-order graph neural networks. In AAAI, volume 33, 4602--4609

2019
[36]

Paszke, A.; Gross, S.; Massa, F.; Lerer, A.; Bradbury, J.; Chanan, G.; Killeen, T.; Lin, Z.; Gimelshein, N.; Antiga, L.; et al. 2019. Pytorch: An imperative style, high-performance deep learning library. NeurIPS

2019
[37]

P.; Luu, A

Ramp \'a s ek, L.; Galkin, M.; Dwivedi, V. P.; Luu, A. T.; Wolf, G.; and Beaini, D. 2022. Recipe for a general, powerful, scalable graph transformer. NeurIPS, 35: 14501--14515

2022
[38]

Said, A.; Bayrak, R.; Derr, T.; Shabbir, M.; Moyer, D.; Chang, C.; and Koutsoukos, X. 2023. Neurograph: Benchmarks for graph machine learning in brain connectomics. NeurIPS

2023
[39]

Sankar, A.; Wu, Y.; Gou, L.; Zhang, W.; and Yang, H. 2018. Dynamic graph representation learning via self-attention networks. arXiv preprint arXiv:1812.09430

2018 arXiv
[40]

Shazeer, N.; Mirhoseini, A.; Maziarz, K.; Davis, A.; Le, Q.; Hinton, G.; and Dean, J. 2017. Outrageously large neural networks: The sparsely-gated mixture-of-experts layer. arXiv preprint arXiv:1701.06538

2017 arXiv
[41]

Spearman, C. 1904. The Proof and Measurement of Association between Two Things. The American Journal of Psychology

1904
[42]

Sporns, O. 2011. The human connectome: a complex network. Annals of the new York Academy of Sciences

2011
[43]

Tan, Z.; Ding, K.; Guo, R.; and Liu, H. 2022 a . Supervised graph contrastive learning for few-shot node classification. In Joint European Conference on Machine Learning and Knowledge Discovery in Databases

2022
[44]

Tan, Z.; Wang , S.; Ding, K.; Li, J.; and Liu, H. 2022 b . Transductive Linear Probing: A Novel Framework for Few-Shot Node Classification. In LoG

2022
[45]

Thiebaut de Schotten, M.; and Forkel, S. J. 2022. The emergent properties of the connected brain. Science

2022
[46]

W.; R \'e , C.; and Poldrack, R

Thomas, A. W.; R \'e , C.; and Poldrack, R. A. 2022. Interpreting mental state decoding with deep learning models. Trends in Cognitive Sciences, 26(11): 972--986

2022
[47]

C.; Smith, S

Van Essen, D. C.; Smith, S. M.; Barch, D. M.; Behrens, T. E.; Yacoub, E.; Ugurbil, K.; Consortium, W.-M. H.; et al. 2013. The WU-Minn human connectome project: an overview. Neuroimage

2013
[48]

N.; Kaiser, .; and Polosukhin, I

Vaswani, A.; Shazeer, N.; Parmar, N.; Uszkoreit, J.; Jones, L.; Gomez, A. N.; Kaiser, .; and Polosukhin, I. 2017. Attention is all you need. NeurIPS, 30

2017
[49]

Veli c kovi \'c , P.; Cucurull, G.; Casanova, A.; Romero, A.; Lio, P.; and Bengio, Y. 2018. Graph attention networks. In ICLR

2018
[50]

Wang, C.; Tsepa, O.; Ma, J.; and Wang, B. 2024 a . Graph-mamba: Towards long-range graph sequence modeling with selective state spaces. arXiv preprint arXiv:2402.00789

2024 arXiv
[51]

J.; and Hsieh, C.-J

Wang, R.; An, S.; Cheng, M.; Zhou, T.; Hwang, S. J.; and Hsieh, C.-J. 2024 b . One Prompt is not Enough: Automated Construction of a Mixture-of-Expert Prompts. arXiv preprint arXiv:2407.00256

2024 arXiv
[52]

Wang , S.; Chen, C.; and Li, J. 2022. Graph Few-shot Learning with Task-specific Structures. In NeurIPS

2022
[53]

Wang , S.; Chen, Z.; Shi, C.; Shen, C.; and Li, J. 2024. Mixture of Demonstrations for In-Context Learning. In NeurIPS

2024
[54]

Wang , S.; Tan, Z.; Liu, H.; and Li, J. 2023. Contrastive Meta-Learning for Few-shot Node Classification. In SIGKDD

2023
[55]

Wang, X.; Yao, L.; Rekik, I.; and Zhang, Y. 2022. Contrastive functional connectivity graph learning for population-based fMRI classification. In International Conference on Medical Image Computing and Computer-Assisted Intervention, 221--230. Springer

2022
[56]

Xu, Y.; Peng, Z.; Shi, B.; Hua, X.; and Dong, B. 2024. Learning dynamic graph representations through timespan view contrasts. Neural Networks, 176: 106384

2024
[57]

Xu, Y.; Shi, B.; Ma, T.; Dong, B.; Zhou, H.; and Zheng, Q. 2023. CLDG: Contrastive learning on dynamic graphs. In 2023 IEEE 39th International Conference on Data Engineering (ICDE), 696--707. IEEE

2023
[58]

Yang, Y.; Zhu, Y.; Cui, H.; Kan, X.; He, L.; Guo, Y.; and Yang, C. 2022. Data-efficient brain connectome analysis via multi-task meta-learning. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 4743--4751

2022
[59]

Ying, Z.; Bourgeois, D.; You, J.; Zitnik, M.; and Leskovec, J. 2019. Gnnexplainer: Generating explanations for graph neural networks. NeurIPS

2019
[60]

You, Y.; Chen, T.; Sui, Y.; Chen, T.; Wang, Z.; and Shen, Y. 2020. Graph contrastive learning with augmentations. NeurIPS

2020
[61]

Zhang, X.; Wang, S.; Lin, N.; Zhang, J.; and Zong, C. 2022. Probing word syntactic representations in the brain by a feature elimination method. In AAAI, volume 36, 11721--11729

2022
[62]

Zhang, Z.; Ji, J.; and Liu, J. 2024. MetaRLEC: Meta-Reinforcement Learning for Discovery of Brain Effective Connectivity. In AAAI

2024
[63]

Zhou, J.; Cui, G.; Hu, S.; Zhang, Z.; Yang, C.; Liu, Z.; Wang, L.; Li, C.; and Sun, M. 2020. Graph neural networks: A review of methods and applications. AI open, 1: 57--81

2020

Pith tools

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

[1] [1]

, " * write output.state after.block = add.period write newline

ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

[2] [2]

write newline

" write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

[3] [3]

S.; and Sporns, O

Bassett, D. S.; and Sporns, O. 2017. Network neuroscience. Nature neuroscience, 20(3): 353--364

work page 2017

[4] [4]

Chang, C.-H.; Lin, C.-H.; and Lane, H.-Y. 2021. Machine learning and novel biomarkers for the diagnosis of Alzheimer’s disease. International journal of molecular sciences, 22(5): 2761

work page 2021

[5] [5]

Chen, J.; Qi, Y.; Wang, Y.; and Pan, G. 2024. Bridging the Semantic Latent Space between Brain and Machine: Similarity Is All You Need. AAAI, 38: 11302--11310

work page 2024

[6] [6]

Cui, H.; Dai, W.; Zhu, Y.; Kan, X.; Gu, A. A. C.; Lukemire, J.; Zhan, L.; He, L.; Guo, Y.; and Yang, C. 2022 a . Braingb: a benchmark for brain network analysis with graph neural networks. IEEE transactions on medical imaging

work page 2022

[7] [7]

Cui, H.; Lu, Z.; Li, P.; and Yang, C. 2022 b . On positional and structural node features for graph neural networks on non-attributed graphs. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management, 3898--3902

work page 2022

[8] [8]

Z.; Rueckert, D.; and Robinson, E

Dahan, S.; Williams, L. Z.; Rueckert, D.; and Robinson, E. C. 2021. Improving phenotype prediction using long-range spatio-temporal dynamics of functional connectivity. In Machine Learning in Clinical Neuroimaging: 4th International Workshop

work page 2021