SYNTOLOGY HomeExplorerAtlasCodeMethodologyAboutDevelopersFeedPricing
Paper · 2608.23936 · 2026

MnemoDyn: Learning Resting State Dynamics from 40K FMRI sequences

Vikas Singh, Won Kim, Tingting Dan, Guorong Wu, Sourav Pal, Viet Luong, Hoseok Lee, Richard Davidson

arXiv · PDF · Open in the Atlas

Code that ran

6 of 8 samples (functions and classes) lifted from this paper's linked repositories executed on a synthesized input in Syntology's jail (0 of them constructed an object rather than computing a result; 1 honoured the contract we drafted, 0 violated it and 0 have no contract verdict; 5 failures of our instrument; 2 no recorded run), measured 2026-11-22; bodies inline only under a permissive licence as GitHub detects it. Not a reproduction of the paper's results.

"Ran" means the function executed on a synthesized input and returned a value. It is not a reproduction of the paper's results. This record, dated, as JSON: /api/ran/2608.23936.json · badge: /api/ran/2608.23936.svg

Each repository's role is a label: a classifier's output recorded on the link, or, where no role is recorded, what the Papers with Code archive marked. The line under it names its basis. Neither says who wrote the code.

RepositoryRoleRan
vsingh-group/mnemodyn canonical (classifier label)
classifier basis: census
6 ran · 2 no recorded run
FunctionStatusWhere it lives
compute_acc Ran
ran; our contract draft was wrong, not the code
vsingh-group/mnemodyn/code/light/abide_classification.py
code served (permissive licence) · get_code("76ae6692cf6c83ff")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned Tensor, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# Create small tensors for pred and target
# pred should have shape (batch_size, num_classes) for softmax
# target should have shape (batch_size,) with class indices

batch_size = 2
num_classes = 4

pred = torch.randn(batch_size, num_classes)
target = torch.tensor([0, 2])

ARGS = {
    "pred": pred,
    "target": target
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type Tensor, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
get_memory Ran
ran, honoured the contract we drafted
vsingh-group/mnemodyn/code/light/abide_classification.py
code served (permissive licence) · get_code("ae078e97ac146b18")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned float, ndim 0, dtype family float, shape () as a later run recorded it. The image it ran in was not recorded. The code mentions cuda, and the harness aliases cuda to the CPU for every sample; whether this run needed that is not recorded. If your copy returns the same type, ndim and shape on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import torch

torch.manual_seed(0)

# The function uses device.type to check if it's 'cuda'
# We'll test with a CPU device
device = torch.device('cpu')

# Allocate some memory on CPU to test the function
# (though it will return nan for non-cuda devices)
x = torch.randn(2, 4)

ARGS = {
    'device': device,
    'reset': False,
    'in_mb': True,
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
torch — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type float, ndim 0, dtype family float, shape () — shape () as a later run recorded it
dtype family
int means an integer dtype; float means any other dtype the instrument saw, bool and complex included (fixture_check.py records only these two families)
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
load_abide_file_list Ran
ran; our contract draft was wrong, not the code
vsingh-group/mnemodyn/code/light/abide_dataset.py
code served (permissive licence) · get_code("0a6623d0572fbcb1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
import tempfile
import pandas as pd
from typing import List, Dict, Tuple

# Create a temporary directory for test data
temp_dir = tempfile.mkdtemp()

# Create test data files
test_file_1 = os.path.join(temp_dir, "test_file_1.dtseries.nii")
test_file_2 = os.path.join(temp_dir, "test_file_2.dtseries.nii")

with open(test_file_1, 'w') as f:
    f.write("dummy data 1")

with open(test_file_2, 'w') as f:
    f.write("dummy data 2")

# Create metadata CSV
metadata_path = os.path.join(temp_dir, "metadata.csv")
metadata_df = pd.DataFrame({
    'relative_path': ['test_file_1.dtseries.nii', 'test_file_2.dtseries.nii'],
    'label_num': [1, 0],
    'gender': [0, 1],
    'age': [25.5, 30.2],
    'dx': [1, 2]
})
metadata_df.to_csv(metadata_path, index=False)

ARGS = {
    'data_dir': temp_dir,
    'metadata_path': metadata_path,
    'label_column': 'dx'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
load_adni_amyloid_file_list Ran
ran; our contract draft was wrong, not the code
vsingh-group/mnemodyn/code/light/adni_dataset.py
code served (permissive licence) · get_code("7fe6252c2f269746")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import pandas as pd
from pathlib import Path
import tempfile
import os

# Create a temporary directory structure with dummy files
tmpdir = tempfile.mkdtemp()
dtseries_root = tmpdir

# Create some dummy dtseries files
os.makedirs(os.path.join(dtseries_root, "subj001"), exist_ok=True)
os.makedirs(os.path.join(dtseries_root, "subj002"), exist_ok=True)

# Create dummy nifti files with the expected naming pattern
Path(os.path.join(dtseries_root, "subj001", "scan_450_parcellated.dtseries.nii")).touch()
Path(os.path.join(dtseries_root, "subj002", "scan_450_parcellated.dtseries.nii")).touch()

# Create the CSV file
csv_data = {
    'PTID': [
        'subj001/scan_parcellated.dtseries.nii',
        'subj002/scan_parcellated.dtseries.nii'
    ],
    'AV45': [1.2, 0.8],
    'LABEL': ['positive', 'negative']
}

df = pd.DataFrame(csv_data)
amyloid_csv = os.path.join(tmpdir, "amyloid.csv")
df.to_csv(amyloid_csv, index=False)

ARGS = {
    'dtseries_root': dtseries_root,
    'amyloid_csv': amyloid_csv
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
load_adni_file_list Ran
ran; our contract draft was wrong, not the code
vsingh-group/mnemodyn/code/light/adni_dataset.py
code served (permissive licence) · get_code("ba69a8c45c8d21c1")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned tuple. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
import pandas as pd
import tempfile
import shutil

# Create a temporary directory for test data
temp_dir = tempfile.mkdtemp()
data_dir = temp_dir

# Create a subdirectory structure for test files
test_subdir = os.path.join(data_dir, "subj001", "scans")
os.makedirs(test_subdir, exist_ok=True)

# Create dummy .dtseries.nii files
dummy_file_1 = os.path.join(test_subdir, "bold_450_parcellated.dtseries.nii")
dummy_file_2 = os.path.join(test_subdir, "bold2_450_parcellated.dtseries.nii")
with open(dummy_file_1, 'w') as f:
    f.write("dummy nifti data")
with open(dummy_file_2, 'w') as f:
    f.write("dummy nifti data")

# Create metadata CSV
metadata_df = pd.DataFrame({
    'relative_path': [
        'subj001/scans/bold_parcellated.dtseries.nii',
        'subj001/scans/bold2_parcellated.dtseries.nii',
    ],
    'researchGroup': ['CN', 'MCI'],
    'subjectAge': [65.5, 72.3],
    'subjectSex': ['M', 'F']
})

metadata_path = os.path.join(temp_dir, 'metadata.csv')
metadata_df.to_csv(metadata_path, index=False)

ARGS = {
    'data_dir': data_dir,
    'metadata_path': metadata_path,
    'label_column': 'researchGroup'
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
pandas, tqdm — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type tuple — a tuple output has no array shape of its own; none is stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
setup_logging Ran
ran; our contract draft was wrong, not the code
vsingh-group/mnemodyn/code/light/abide_classification.py
code served (permissive licence) · get_code("7ab65340af034176")
How it ran

Ran in Syntology's sandbox on this synthesized input (drafted by a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API) and returned RootLogger. The image it ran in was not recorded. If your copy returns the same type on these inputs, it matched what we observed; nothing here checks that. It does NOT mean the output is correct: nothing has compared it to the paper.

import os
import tempfile
import logging

# Create a temporary directory for logging
log_dir = tempfile.mkdtemp()

# Define ARGS dictionary with all parameters for setup_logging
ARGS = {
    'log_dir': log_dir
}
Setup block drafted by
a language model, Claude Haiku 4.5, prompt python_fixture_v1, through the batch API
Imports
none outside the standard library — import names as written, not package names to install; a name the repository itself provides is not a package
Returned
type RootLogger — a RootLogger output has no array shape of its own; none is stated — the instrument coerces any output to an array; for a RootLogger output the recorded ndim 0 and dtype family float describe that coercion, not the output, and are not stated
Environment
image not recorded; re-verify against syntology-harvest-sandbox:v5-deps; not recorded for this sample: the harness installs a cuda-to-cpu shim (cuda_shim.py) before every sample and no run records whether a sample needed it; seed: none by the harness; any seed is in the setup block
load_adhd_file_list Not yet run
unverified (harvested, no recorded run)
vsingh-group/mnemodyn/code/light/adhd_dataset.py
code served (permissive licence) · get_code("c99f508184e2a1d5")
How it ran

no invocation recorded.

load_hbn_file_list Not yet run
unverified (harvested, no recorded run)
vsingh-group/mnemodyn/code/light/hbn_dataset.py
code served (permissive licence) · get_code("e00b4682f9a29561")
How it ran

no invocation recorded.

Also present in other papers' code

Of 8 samples lifted from this paper's linked repositories, 8 carry a function identity and 0 do not. Of the 8 distinct functions with an identity, 0 are also present in other papers whose code we hold (0 with at least one paper dated earlier, 0 the same month, 0 later) and 8 are only here among the 52,977 papers whose code Syntology holds (as of 2026-09-26).

Only here means: among the 52,977 papers whose code Syntology holds (51,887 with lifted samples, 1,090 through their assemblies only), as of 2026-09-26, no other paper's lifted samples or current assembly bricks contain an identical function body after normalisation.

Presence, not direction: an identical function body in two papers' repositories does not say who wrote it, which came from which, or that either paper's method rests on it. Matching is exact after normalisation, so a renamed or edited function is not matched. Dates are when papers appeared, not when code was written. Only .py files the harvester lifted are covered; notebooks are not read. A new-style arXiv id YYMM.NNNN dates a paper to the month 20YY-MM; otherwise Paper.year dates it to the year; otherwise it is undated. Dates are when the paper appeared, not when the code was written. Read 2026-11-22.

Repositories linked to this paper

Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.

Abstract

We present a dynamical-systems based model for resting-state functional magnetic resonance imaging (rs-fMRI), trained on a dataset of roughly 40K rs-fMRI sequences covering a wide variety of public and available-by-permission datasets. While most existing proposals use transformer backbones, we utilize multi-resolution temporal modeling of the dynamics across parcellated brain regions. We show that MnemoDyn is compute efficient and generalizes very well across diverse populations and scanning protocols. When benchmarked against current state-of-the-art transformer-based approaches, MnemoDyn consistently delivers superior reconstruction quality. Overall, we find that with such large-scale pre-training on (non-proprietary) rs-fMRI datasets, we get a highly performant model for various downstream tasks. Our results also provide evidence of the efficacy of the model on small sample size studies which has implications for neuroimaging studies at large where resting state fMRI is a commonly acquired imaging modality.

For agents

The same record, over MCP at https://syntology.ai/mcp:

get_harvested_code_for_paper("2608.23936")
get_code_for_paper("2608.23936")
have("2608.23936")

The run record, dated, one paper per request, free:

curl https://syntology.ai/api/ran/2608.23936.json

A badge for a README (the split and the date, never a ratio):

[![Syntology run record](https://syntology.ai/api/ran/2608.23936.svg)](https://syntology.ai/paper/2608.23936)

Connect an agent — have() is free.