OSCR

MambaKAN: An Interpretable Framework for Alzheimer's Disease Diagnosis via Selective State Space Modeling of Dynamic Functional Connectivity.

Code ↔ Paper

17 matches between paragraphs of the paper and lines of its authors' code, computed by the harvester (lexical-v1). Click a colored paragraph or line to see its counterpart.

The 17 matches
  1. [1] § 3. Methods › 3.5. Joint Training Strategy › 3.5.1. Two-Phase Training Protocol ↔ train_stage1.py, lines 1–30 · score 0.92 · lowest validation reconstruction, VAE Unsupervised Pre, noise robust latent, best checkpoint, pre training, Adam
  2. [2] § 3. Methods › 3.4. Kolmogorov–Arnold Network Classifier › 3.4.1. KAN Architecture Motivation ↔ models/kan_classifier.py, lines 1–17 · score 0.84 · Kolmogorov Arnold Networks, learnable weights, visualizable curve, intrinsically interpretable, inspired, MLPs
  3. [3] § 3. Methods › 3.3. Mamba Selective State Space Temporal Encoder › 3.3.3. Selective State Space Model (S6) ↔ models/mamba_encoder.py, lines 13–22 · score 0.82 · Selective State Space, depthwise convolution, selectivity mechanism, Mamba block, S6, Model
  4. [4] § 3. Methods › 3.5. Joint Training Strategy › 3.5.1. Two-Phase Training Protocol ↔ train_stage2.py, lines 1–44 · score 0.76 · End Joint Fine, trained jointly, MambaKAN, tuning, pipeline, accuracy
  5. [5] § 3. Methods › 3.4. Kolmogorov–Arnold Network Classifier › 3.4.5. MambaKAN Classifier Structure ↔ models/kan_classifier.py, lines 180–280 · score 0.73 · layer KAN, context vector, Mamba encoder, KAN classifier, learnable, weights
  6. [6] § 3. Methods › 3.1. Overview of MambaKAN ↔ train_stage2.py, lines 1–44 · score 0.72 · VAE encoding, MambaKAN, KAN classification, Mamba temporal, pipeline, joint
  7. [7] § 5. Interpretability Analysis › 5.3. Layer 3: Gradient-Based Brain Region Attribution ↔ analysis.py, lines 230–357 · score 0.67 · bar charts, attribution scores, chord, heatmap, connectivity, Gradient
  8. [8] § 3. Methods › 3.5. Joint Training Strategy › 3.5.2. Differential Learning Rates ↔ train_stage2.py, lines 46–74 · score 0.63 · warmup epochs, fine tuning, frozen, trained, KAN, Mamba
  9. [9] § 3. Methods › 3.3. Mamba Selective State Space Temporal Encoder › 3.3.1. Rationale for Mamba over LSTM and Transformer ↔ models/mamba_encoder.py, lines 152–179 · score 0.63 · parallel scan, hardware aware, efficient, matrices, Selective, Mamba
  10. [10] § 3. Methods › 3.4. Kolmogorov–Arnold Network Classifier › 3.4.2. Rationale for KAN over MLP ↔ models/kan_classifier.py, lines 1–17 · score 0.63 · provides intrinsic interpretability, latent dimension, MLP, mapping, linear, logits
  11. [11] § 5. Interpretability Analysis › 5.3. Layer 3: Gradient-Based Brain Region Attribution ↔ analysis.py, lines 230–357 · score 0.59 · chord diagram, attribution scores, connectivity, Gradient, Brain, class
  12. [12] § 3. Methods › 3.4. Kolmogorov–Arnold Network Classifier › 3.4.3. B-Spline Edge Activations ↔ models/kan_classifier.py, lines 29–66 · score 0.59 · spline coefficients, spline basis functions, uniform, learnable, weight
  13. [13] § 3. Methods › 3.3. Mamba Selective State Space Temporal Encoder › 3.3.5. Temporal Context Aggregation ↔ models/mamba_encoder.py, lines 182–235 · score 0.58 · stacked Mamba blocks, temporal context, sequence, vector
  14. [14] § 5. Interpretability Analysis › 5.3. Layer 3: Gradient-Based Brain Region Attribution ↔ analysis.py, lines 1–41 · score 0.55 · attribution matrix, pairwise, ROI, brain, map, Gradient
  15. [15] § 3. Methods › 3.2. Variational Autoencoder for Per-Window Feature Extraction › 3.2.3. VAE Loss Function ↔ models/vae.py, lines 51–55 · score 0.54 · KL divergence, reconstruction loss, VAE
  16. [16] § 3. Methods › 3.5. Joint Training Strategy › 3.5.3. Regularization ↔ models/mamba_encoder.py, lines 13–22 · score 0.54 · depthwise convolution, Mamba block, Dropout, space
  17. [17] § 4. Experiments › 4.2. Evaluation Metrics ↔ analysis.py, lines 1–41 · score 0.52 · ROC curve, MambaKAN, AUC

Paper

Loaded from Europe PMC by your browser, not stored by OSCR: doi.org · Europe PMC

The paper is loaded when this pane is shown.

The authors' code

Python · 297 lines · 10 KB · no license · 4 matches

  1. """
  2. Kolmogorov-Arnold Network (KAN) Classifier — B-Spline implementation.
  4. Reference: Liu et al., "KAN: Kolmogorov-Arnold Networks" (2024)
  5. Inspired by efficient-kan (https://github.com/Blealtan/efficient-kan)
  7. Key difference from MLP:
  8. MLP: y = W · σ(x) — fixed activation, learnable weights
  9. KAN: y = Σ φ_{q,p}(x_p) — learnable spline activations per connection
  11. Visualizing φ curves directly reveals the non-linear mapping from each
  12. latent dimension to class logits, providing intrinsic interpretability.
  13. """
  14. import math
  15. import torch
  16. import torch.nn as nn
  17. import torch.nn.functional as F
  20. class KANLinear(nn.Module):
  21. """
  22. A single KAN layer: replaces Linear + fixed activation with
  23. per-connection learnable B-spline activations.
  25. Input: (batch, in_features)
  26. Output: (batch, out_features)
  27. """
  29. def __init__(
  30. self,
  31. in_features: int,
  32. out_features: int,
  33. grid_size: int = 5,
  34. spline_order: int = 3,
  35. scale_noise: float = 0.1,
  36. scale_base: float = 1.0,
  37. grid_range: tuple = (-1.0, 1.0),
  38. ):
  39. super().__init__()
  40. self.in_features = in_features
  41. self.out_features = out_features
  42. self.grid_size = grid_size
  43. self.spline_order = spline_order
  45. # Build extended B-spline grid
  46. h = (grid_range[1] - grid_range[0]) / grid_size
  47. grid = torch.linspace(
  48. grid_range[0] - spline_order * h,
  49. grid_range[1] + spline_order * h,
  50. grid_size + 2 * spline_order + 1,
  51. )
  52. self.register_buffer("grid", grid)
  54. # Number of B-spline basis functions
  55. n_basis = grid_size + spline_order
  57. self.base_weight = nn.Parameter(torch.empty(out_features, in_features))
  58. # Spline coefficients: (out, in, n_basis)
  59. self.spline_weight = nn.Parameter(torch.empty(out_features, in_features, n_basis))
  61. # Per-connection scaling factors (learnable)
  62. self.scale_base = nn.Parameter(torch.ones(out_features, in_features) * scale_base)
  63. self.scale_spline = nn.Parameter(torch.ones(out_features, in_features))
  65. nn.init.kaiming_uniform_(self.base_weight, a=math.sqrt(5))
  66. nn.init.normal_(self.spline_weight, mean=0.0, std=scale_noise)
  68. def b_splines(self, x: torch.Tensor) -> torch.Tensor:
  69. """
  70. Evaluate B-spline basis functions at x via Cox–de Boor recursion.
  72. Args:
  73. x: (batch, in_features)
  74. Returns:
  75. bases: (batch, in_features, n_basis)
  76. """
  77. assert x.dim() == 2
  78. x = x.unsqueeze(-1)
  79. grid = self.grid
  81. # Order-0 indicator basis
  82. bases = ((x >= grid[:-1]) & (x < grid[1:])).float()
  84. # Cox–de Boor recursion
  85. for k in range(1, self.spline_order + 1):
  86. denom_l = grid[k:-1] - grid[: -(k + 1)]
  87. denom_r = grid[k + 1 :] - grid[1: -k]
  88. # Avoid division by zero
  89. left = torch.where(
  90. denom_l != 0,
  91. (x - grid[: -(k + 1)]) / denom_l * bases[..., :-1],
  92. torch.zeros_like(bases[..., :-1]),
  93. )
  94. right = torch.where(
  95. denom_r != 0,
  96. (grid[k + 1 :] - x) / denom_r * bases[..., 1:],
  97. torch.zeros_like(bases[..., 1:]),
  98. )
  99. bases = left + right
  101. return bases # (b, in, n_basis)
  103. def forward(self, x: torch.Tensor) -> torch.Tensor:
  104. """
  105. Args:
  106. x: (batch, in_features)
  107. Returns:
  108. out: (batch, out_features)
  109. """
  110. # Base (SiLU) branch
  111. base_out = F.linear(F.silu(x), self.base_weight * self.scale_base)
  113. # Spline branch: contract (b, in, n_basis) × (out, in, n_basis) → (b, out)
  114. bases = self.b_splines(x)
  115. spline_out = torch.einsum(
  116. "bik,oik->bo",
  117. bases,
  118. self.spline_weight * self.scale_spline.unsqueeze(-1),
  119. )
  121. return base_out + spline_out
  123. def get_activation_curve(self, dim: int, n_points: int = 200, x_range=(-3.0, 3.0)):
  124. """
  125. Evaluate the learned spline activation φ(x) for a single input dimension.
  127. Args:
  128. dim: input dimension index
  129. n_points: number of evaluation points
  130. x_range: evaluation range
  131. Returns:
  132. x_vals: (n_points,)
  133. y_vals: (out_features, n_points)
  134. """
  135. x_vals = torch.linspace(x_range[0], x_range[1], n_points, device=self.grid.device)
  136. dummy = torch.zeros(n_points, self.in_features, device=self.grid.device)
  137. dummy[:, dim] = x_vals
  139. bases = self.b_splines(dummy)
  140. sw = self.spline_weight[:, dim, :] * self.scale_spline[:, dim].unsqueeze(-1)
  141. y_vals = (bases[:, dim, :] @ sw.T).T # (out, n_pts)
  142. return x_vals.detach(), y_vals.detach()
  144. def update_grid(self, x: torch.Tensor, margin: float = 0.01):
  145. """
  146. Adapt the B-spline grid to span the activation range of x.
  148. Note on coefficient resampling:
  149. A full grid update should re-fit spline_weight onto the new basis
  150. (e.g. via least squares) to preserve learned activation shapes.
  151. This implementation omits that step because the current architecture
  152. uses a single shared 1-D grid for all in_features; when the grid
  153. shifts significantly the new and old basis spaces diverge and simple
  154. least-squares resampling is numerically unreliable (larger error
  155. than no resampling at all on large feature dimensions).
  156. In practice, call this method infrequently (every 50 epochs) so the
  157. network has enough gradient steps to recover from the small
  158. coefficient mismatch before the next update.
  160. Args:
  161. x: (batch, in_features) — representative input batch
  162. margin: fractional padding added beyond [x_min, x_max]
  163. """
  164. with torch.no_grad():
  165. x_min, x_max = x.min().item(), x.max().item()
  166. span = max(x_max - x_min, 1e-6)
  167. x_min -= margin * span
  168. x_max += margin * span
  169. h = (x_max - x_min) / self.grid_size
  170. new_grid = torch.linspace(
  171. x_min - self.spline_order * h,
  172. x_max + self.spline_order * h,
  173. len(self.grid),
  174. device=self.grid.device,
  175. dtype=self.grid.dtype,
  176. )
  177. self.grid.copy_(new_grid)
  180. class KANClassifier(nn.Module):
  181. """
  182. Two-layer KAN for classification.
  184. Structure: in_features → hidden_dim → num_classes
  186. Parameter count note:
  187. This implementation includes per-connection learnable scaling factors
  188. (scale_base and scale_spline in each KANLinear layer) for training
  189. stability. These add (out * in) extra parameters per layer compared
  190. to counting only the spline and base weights. As a result the KAN
  191. head has ~93 K parameters, somewhat larger than the ~43 K figure
  192. reported in the paper (which counted only spline_weight + base_weight).
  193. """
  195. def __init__(
  196. self,
  197. in_features: int,
  198. hidden_dim: int,
  199. num_classes: int,
  200. grid_size: int = 5,
  201. spline_order: int = 3,
  202. ):
  203. super().__init__()
  204. self.layer1 = KANLinear(in_features, hidden_dim, grid_size, spline_order)
  205. self.layer2 = KANLinear(hidden_dim, num_classes, grid_size, spline_order)
  207. def forward(self, x: torch.Tensor) -> torch.Tensor:
  208. return self.layer2(self.layer1(x))
  210. def update_grid(self, x: torch.Tensor):
  211. """
  212. Update B-spline grids in both layers.
  213. Call with a representative context batch (output of MambaEncoder)
  214. every few epochs during Phase-2 training.
  216. Args:
  217. x: (batch, in_features) — Mamba context vectors
  218. """
  219. self.layer1.update_grid(x)
  220. with torch.no_grad():
  221. h = self.layer1(x)
  222. self.layer2.update_grid(h)
  224. def get_input_importance(self) -> torch.Tensor:
  225. """
  226. L1 norm of spline weights in layer 1, summed over outputs and basis functions.
  228. Returns:
  229. importance: (in_features,) — higher = more influential
  230. """
  231. w = self.layer1.spline_weight # (out, in, n_basis)
  232. return w.abs().sum(dim=[0, 2]) # (in_features,)
  234. def get_class_curves(
  235. self,
  236. dim: int,
  237. x_mean: torch.Tensor,
  238. n_points: int = 200,
  239. x_range: tuple = (-3.0, 3.0),
  240. ):
  241. """
  242. Ceteris-paribus class activation curve for input dimension `dim`.
  243. Varies dim over x_range while holding all other dims at x_mean.
  245. Args:
  246. dim: input dimension to vary
  247. x_mean: (in_features,) anchor values for non-varied dims
  248. n_points: evaluation resolution
  249. x_range: range of variation for `dim`
  250. Returns:
  251. x_vals: (n_points,)
  252. logits: (num_classes, n_points)
  253. """
  254. device = next(self.parameters()).device
  255. x_vals = torch.linspace(x_range[0], x_range[1], n_points, device=device)
  256. probe = x_mean.unsqueeze(0).expand(n_points, -1).clone().to(device)
  257. probe[:, dim] = x_vals
  258. with torch.no_grad():
  259. out = self.forward(probe) # (n_points, num_classes)
  260. return x_vals.cpu(), out.T.cpu() # (num_classes, n_points)
  262. def get_top_activation_curves(self, top_k: int = 10, n_points: int = 200):
  263. """
  264. Returns activation curves for the top-k most important input dimensions.
  266. Returns:
  267. top_dims: (top_k,)
  268. x_vals: (n_points,)
  269. curves: (top_k, out_features, n_points)
  270. """
  271. importance = self.get_input_importance()
  272. top_dims = importance.topk(top_k).indices
  274. curves, x_vals = [], None
  275. for d in top_dims.tolist():
  276. xv, yv = self.layer1.get_activation_curve(d, n_points)
  277. x_vals = xv
  278. curves.append(yv)
  280. return top_dims, x_vals, torch.stack(curves, dim=0)
  283. if __name__ == "__main__":
  284. torch.manual_seed(0)
  285. b, in_f, hidden, num_cls = 16, 128, 64, 4
  287. model = KANClassifier(in_f, hidden, num_cls)
  288. x = torch.randn(b, in_f)
  289. logits = model(x)
  290. print("Logits shape:", logits.shape) # (16, 4)
  292. imp = model.get_input_importance()
  293. print("Importance shape:", imp.shape) # (128,)
  295. top_dims, x_vals, curves = model.get_top_activation_curves(top_k=5)
  296. print("Top dims:", top_dims.tolist())
  297. print("Curves shape:", curves.shape) # (5, 64, 200)

kan_classifier.py at commit 0ff30c0, no license · at the source

Overview

Authors: Libin Gao1, Zhongyi Hu2
ORCID iDs: Libin Gao, Zhongyi Hu
  1. Artificial Intelligence College, Zhejiang Industry & Trade Vocational College, Wenzhou 325000, China
  2. College of Computer Science and Artificial Intelligence, Wenzhou University, Wenzhou 325035, China
Institutions: Zhejiang Industry & Trade Vocational College (China); Wenzhou University (China)
Journal: Brain sciences, volume 16, issue 4, article 421
Dates: received 25 March 2026; accepted 16 April 2026; published online 17 April 2026
Type: Research article · Language: English
License: CC BY
Identifiers: DOI 10.3390/brainsci16040421 · PMID 42041829 · PMCID PMC13114612 · OpenAlex W7154712876
Open access: gold, a free copy (OpenAlex)
Status: code verified
Categories: fMRI (modality), Alzheimer's / dementia (population)
Methods: Spectral & time-frequency, Statistics, Preprocessing, Connectivity, Machine learning, fMRI & imaging
Keywords: Alzheimer’s disease, dynamic functional connectivity, state space model, Mamba, Kolmogorov–Arnold network, variational autoencoder, interpretability, resting-state fMRI, deep learning, brain connectivity
Topic: Functional Brain Connectivity Studies (Cognitive Neuroscience, Neuroscience), according to OpenAlex
Funding: Wenzhou major scientific and technological innovation project plan “reveal the list” (ZG2023022); Wenzhou Fundamental Scientific Research Program (Y20240185, G20240065); General Program of Wenzhou Municipal Science and Technology Bureau (S2023013); General Program of Zhejiang Provincial Department of Education (Y202250720); Teacher Scientific Research Project of Zhejiang Industry & Trade Vocational College (G250101)
Citations: cited by 1 paper (Europe PMC); 24 references in the paper

Abstract

Background/Objectives: Alzheimer’s disease (AD) is an irreversible neurodegenerative disorder that imposes a profound burden on global public health. While resting-state functional magnetic resonance imaging (rs-fMRI)-based dynamic functional connectivity (dFC) analysis has demonstrated promise in capturing time-varying brain network abnormalities, existing deep learning methods suffer from three fundamental limitations: (1) an inability to model temporal dependencies across dynamic connectivity windows, (2) reliance on post hoc black-box explainability tools, and (3) misalignment between feature learning and classification objectives. Methods: To address these challenges, we propose MambaKAN, an end-to-end interpretable framework integrating a Variational Autoencoder (VAE), a Selective State Space Model (Mamba), and a Kolmogorov–Arnold Network (KAN). The VAE encodes each dFC snapshot into a compact latent representation, preserving nonlinear connectivity patterns. The Mamba encoder captures long-range temporal dynamics across the sequence of latent representations via input-selective state transitions. The KAN classifier provides intrinsic interpretability through learnable B-spline activation functions, enabling direct visualization of how latent features influence diagnostic decisions without post-hoc approximation. The entire pipeline is trained end-to-end with a joint loss function that aligns feature learning with classification. Results: Evaluated on the Alzheimer’s Disease Neuroimaging Initiative (ADNI) dataset across five classification tasks (CN vs. AD, CN vs. EMCI, EMCI vs. LMCI, LMCI vs. AD, and four-class), MambaKAN achieves accuracies of 95.1%, 89.8%, 84.0%, 86.7%, and 70.5%, respectively, outperforming strong baselines including LSTM, Transformer, and MLP-based variants. Conclusions: Comprehensive ablation studies confirm the indispensable contribution of each module, and the three-layer interpretability analysis reveals key temporal patterns and brain regions associated with AD progression.

Reproduced under the paper's license (CC BY), from the paper cited above.

Repository

Its files are read in the Code ↔ Paper reader above, with 17 matches between paragraphs and lines of code.

l1binn/MambaKAN

License: none: the authors keep all their rights
State: the link answers, verified on 29 September 2026
Evidence: files inventoried
Commit: 0ff30c0a3602f29e7eb6b90543f4ef7978b34707, 17 April 2026
Languages: Python (9)
Size: 11 files, 9 scripts
Software Heritage: not archived
Found in: “Data Availability Statement”
Holds: README
Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation
Tools: PyTorch (8 files), Matplotlib (1 file), NumPy (1 file), scikit-learn (1 file), SciPy (1 file)
Availability: 1 check, the latest on 29 September 2026: the link answers
  • 29 September 2026: the link answers
10 files
At the source: github.com/l1binn/MambaKAN

The paper's code and data availability statement is in the Data section.

Tracing map

Proposed by the machine: these links were found in the paper and verified at the source, without human review. The map will receive a Zenodo DOI once one of the paper's authors has validated it with their ORCID.

What the map holds:

  • 1 repository of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
  • 9 scripts, each with its path and the digest of its content;
  • 17 matches between paragraphs of the paper and lines of the code (method lexical-v1);
  • neither the text of the paper nor the code itself.

Its JSON (tracing-map.json) is deposited on Zenodo with its DOI once the map is validated.

Data

No dataset and no data link were found in the paper.

Data Availability Statement

The ADNI dataset is publicly available at https://adni.loni.usc.edu/ (accessed on 14 April 2026) upon registration. The core model and algorithm code is publicly available at https://github.com/l1binn/MambaKAN.

Reproduced under the paper's license (CC BY), from the paper cited above.

Versions

The history of this record: each version stored by the harvester or made by a correction of its authors or of the maintainers of its code, and what changed in its facts. The texts of the paper (its abstract, its availability statements) are not part of it; versions that changed only those are not listed.

Version 1, 29 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 2 authors, 10 keywords, 5 funders, 20 references.

Cite

This paper

Gao, L., & Hu, Z. (2026). MambaKAN: An Interpretable Framework for Alzheimer's Disease Diagnosis via Selective State Space Modeling of Dynamic Functional Connectivity. Brain sciences, 16(4), 421. https://doi.org/10.3390/brainsci16040421

BibTeX

@article{gao2026mambakan,
author = {Gao, Libin and Hu, Zhongyi},
title = {{MambaKAN: An Interpretable Framework for Alzheimer's Disease Diagnosis via Selective State Space Modeling of Dynamic Functional Connectivity}},
journal = {Brain sciences},
year = {2026},
month = apr,
volume = {16},
number = {4},
pages = {421},
publisher = {Multidisciplinary Digital Publishing Institute (MDPI)},
issn = {2076-3425},
doi = {10.3390/brainsci16040421},
url = {https://doi.org/10.3390/brainsci16040421},
pmid = {42041829},
pmcid = {PMC13114612}
}

RIS

TY - JOUR
AU - Gao, Libin
AU - Hu, Zhongyi
TI - MambaKAN: An Interpretable Framework for Alzheimer's Disease Diagnosis via Selective State Space Modeling of Dynamic Functional Connectivity
T2 - Brain sciences
J2 - Brain Sci
PY - 2026
DA - 2026/04/17
VL - 16
IS - 4
SP - 421
SN - 2076-3425
PB - Multidisciplinary Digital Publishing Institute (MDPI)
DO - 10.3390/brainsci16040421
UR - https://doi.org/10.3390/brainsci16040421
LA - en
ER -

CSL-JSON

{
"id": "10.3390/brainsci16040421",
"type": "article-journal",
"title": "MambaKAN: An Interpretable Framework for Alzheimer's Disease Diagnosis via Selective State Space Modeling of Dynamic Functional Connectivity",
"container-title": "Brain sciences",
"author": [
{
"family": "Gao",
"given": "Libin"
},
{
"family": "Hu",
"given": "Zhongyi"
}
],
"container-title-short": "Brain Sci",
"volume": "16",
"issue": "4",
"page": "421",
"DOI": "10.3390/brainsci16040421",
"PMID": "42041829",
"PMCID": "PMC13114612",
"ISSN": "2076-3425",
"publisher": "Multidisciplinary Digital Publishing Institute (MDPI)",
"URL": "https://doi.org/10.3390/brainsci16040421",
"language": "en",
"issued": {
"date-parts": [
[
2026,
4,
17
]
]
}
}

The tracing map gets a citation of its own once an author has validated it and it has a DOI.

Similar papers

The papers with a page that share the most with this one: the tools found in their code, their categories, datasets, cited references and authors, the rarest counting most.

[1] doi:10.1002/hbm.70483 [code]
Untamed: Unconstrained Tensor Decomposition and Graph Node Embedding for Cortical Parcellation.
Journal: Human brain mapping
In common: PyTorch, scikit-learn, SciPy, 2 other tools, fMRI, 1 reference
[2] doi:10.1038/s41467-026-73687-9 [code]
Automatic selection of the best neural architecture for time series forecasting.
Journal: Nature communications
In common: PyTorch, SciPy, Matplotlib, 1 other tool, 2 references
[3] doi:10.1002/alz.71365 [code]
Benchmarking speech biomarkers of Alzheimer's against cognitive and neural measures.
Journal: Alzheimer's & dementia : the journal of the Alzheimer's Association
In common: scikit-learn, SciPy, Matplotlib, 1 other tool, fMRI, Alzheimer's / dementia, 1 reference
[4] doi:10.1038/s41586-026-10528-1 [code]
A critical initialization for biological neural networks.
Journal: Nature
In common: PyTorch, scikit-learn, SciPy, 2 other tools, 1 reference
[5] doi:10.1162/imag.a.1220 [code]
Brain functional network connectivity interpolation characterizes the neuropsychiatric continuum and heterogeneity.
Journal: Imaging neuroscience (Cambridge, Mass.)
In common: PyTorch, scikit-learn, SciPy, 2 other tools, 1 reference
[6] doi:10.1038/s41598-026-68186-2 [code]
NeuroStream: spectral-spatio-temporal deep learning for visual stimulus classification from EEG.
Journal: Scientific reports
In common: PyTorch, scikit-learn, SciPy, 2 other tools, 1 reference
[7] doi:10.1038/s43856-026-01817-x [code]
Visual prompt engineering for multimodal and irregularly sampled medical data.
Journal: Communications medicine
In common: PyTorch, scikit-learn, SciPy, 2 other tools, 1 reference
[8] doi:10.1098/rstb.2024.0461 [code]
Shallow recurrent decoders for neural and behavioural dynamics.
Journal: Philosophical transactions of the Royal Society of London. Series B, Biological sciences
In common: PyTorch, scikit-learn, SciPy, 2 other tools, 1 reference
[9] doi:10.3390/bioengineering13080924 [code]
Deep Learning-Based Temporal Gait Analysis Using a Smartphone IMU in Older Adults with and Without Non-Specific Low Back Pain.
Journal: Bioengineering (Basel, Switzerland)
In common: PyTorch, scikit-learn, SciPy, 2 other tools, 1 reference
[10] doi:10.1038/s41467-026-75783-2 [code]
Broadband encoding and high-speed probabilistic bit generation with integrated microwave neurons.
Journal: Nature communications
In common: PyTorch, scikit-learn, SciPy, 2 other tools, 1 reference

Contribute

The authors of this paper can claim it, correct its record and validate its tracing map, and the maintainers of its code (its owner, or a public member of its organization) correct what it says of their repository; anyone signed in can ask for its removal. Every request goes to OSCR's own machine, which answers it; your account page follows them.

Sign in with ORCID to claim this paper as one of its authors, correct its record or validate its tracing map: when the paper's metadata lists your ORCID iD, you are recognized at once. Maintainers of its code: sign in with GitHub, then claim the repository on your account page.

Request its removal

To ask OSCR to remove this record, the copies of its authors' scripts or its tracing map, use the removal request page: signed in, you say who you are, what to remove and why, then review and confirm the request. Published rules decide every request (how).

Discussion, reproductions, activity

Discussion: questions and error reports about this paper and its code, from signed-in readers and its authors. It opens with sign-in.

Reproductions: reports from readers who ran the authors' code: what they reproduced, with which environment, commit and data. It opens with sign-in.

Activity: what happens around this paper: new versions of its record, its map's validation, discussions and reproductions. It opens with sign-in.