Population coding under the scale invariance of high-dimensional noise.

Code ↔ Paper

6 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 6 matches

[1] § MATERIALS AND METHODS › Simulation of neural activity with bounded and unbounded LFI ↔ data_simulated_gamma2_manyrep/gen_simdata.py, lines 12–99 · score 0.85 · inverse Wishart distribution, lognormal distribution, firing rates, mouse TX42, Simulation, matrix
[2] § MATERIALS AND METHODS › Bias correcting the eigenvalues and angular occupations › Eigenvalue bias correction ↔ code/code_janelia_trialeffect_new/bias_correct_morevals_allrec.ipynb, lines 12–106 · score 0.58 · initial guess, Bias corrected eigenvalues, iteratively, naive
[3] § MATERIALS AND METHODS › Bias correcting the eigenvalues and angular occupations › Eigenvalue bias correction ↔ code/code_simdata_alphapos_gamma2/bias_correct.ipynb, lines 32–82 · score 0.58 · initial guess, Bias corrected eigenvalues, iteratively, naive
[4] § MATERIALS AND METHODS › Bias correcting the eigenvalues and angular occupations › Angular occupation bias correction › Eigenvector bias correction. ↔ code/code_janelia_trialeffect_new/bias_correct_morevals_allrec.ipynb, lines 12–106 · score 0.55 · initial guess, bias corrected, algorithm, naive, cosines, squared
[5] § MATERIALS AND METHODS › Bias correcting the eigenvalues and angular occupations › Angular occupation bias correction › Eigenvector bias correction. ↔ code/code_simdata_alphapos_gamma2/bias_correct.ipynb, lines 32–82 · score 0.55 · initial guess, bias corrected, algorithm, naive, cosines, squared
[6] § MATERIALS AND METHODS › Analysis of mouse V1 data ↔ code/functions/data_extract.py, lines 26–75 · score 0.51 · 43–47 deg, static, orientations, 43 deg, stimuli

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 paper is loaded by your browser, with JavaScript: read it at doi.org or on Europe PMC.

The authors' code

Jupyter notebook · 109 lines · 4.3 KB · no license · 2 matches

# %%
import sys
sys.path
sys.path.append("..")
sys.path.append("/mnt/disk1/home/sumedh/.local/lib/python3.8/site-packages")
import numpy as np, matplotlib.pyplot as plt, time, math, sklearn
import random
%matplotlib inline
from tqdm.notebook import tqdm, trange
import matplotlib
# %%
from functions.pls_analysis import get_cos2_eig
from functions.bias_correction_functions import bias_correct_eigs, bias_correct_cos2
mouse_ids = ['TX42', 'TX41', 'TX40', 'TX39', 'TX38'] #names of the mice
max_trials = [2654, 2246, 2231, 1999, 1855]
nvals_plot = [150*i for i in range(1, 10)]
# eigenvalue bias correction params
r=100 # Number of max iterations
beta=0.05 # learning rate in log scale
Nsub=200 # Number of subpopulations
#cos2 bias correction params
M=1000 # Number of iterations for cosine bias correction
eta=0.01 # learning rate in log scale
q=10 # number of repeats of the algorithm with different random seed
################################ Loop over all recordings (mice) ################################
biggesttic = time.perf_counter()
for rec_id in [1, 2, 3, 4]:
numtrials_all = [100*i for i in range(5, 26) if 100*i<max_trials[rec_id]] #multiples of 100
T = numtrials_all[-1]
# get values only for specific N indices to perform bias correction
#load results with the largest number of trials
results_i = np.load(f'./results_janelia_trialeffect/rec_{rec_id}/janelia_rec{rec_id}_{T}tr.npz')
cos2, eig = get_cos2_eig(results_i) #averaged over repeats (subsets)
N_vals = results_i['Neuron_subsize']
norm2_delmu = (results_i['norm_delmu']**2).mean(axis=0)
inds_plot = [int(np.argwhere(N_vals==nvals_plot[i])) for i in range(len(nvals_plot))]
eigs_x = eig[:, inds_plot]
cos2_x = cos2[:, inds_plot]
norm2dmu_x = norm2_delmu[inds_plot]
eigs_bc = np.empty((N_vals[inds_plot[-1]], len(inds_plot)))
eigs_bc[:] = np.nan
cos2_bc = np.empty((N_vals[inds_plot[-1]], len(inds_plot)))
cos2_bc[:] = np.nan
bigtic = time.perf_counter()
for i in range(len(inds_plot)):
smalltic = time.perf_counter()
N_curr = N_vals[inds_plot[i]]
w=np.ones(N_curr)*0.1 # Used for smoothing the naive estimates of cosine squares to avoid negative values
#eigenvalue bias correction
eh = eigs_x[:N_curr, i]
ei=eh # initial guess equal to the naive estimate
ebs , cost = bias_correct_eigs(eh,ei,T,r,beta,Nsub)
idx=np.argmin(cost)
eb=ebs[idx,:] # This is the bias corrected eigenvalues
eigs_bc[:N_curr,i] = eb
# print(f"Eig bc for N={N_curr} done in {(time.perf_counter()-smalltic):.2f}s")
with open('status_bc_morevals_allrec.txt', 'a') as f:
f.write(f"\nRec={rec_id}: Eig bc for N={N_curr} done in {(time.perf_counter()-smalltic):.2f}s")
smalltic2 = time.perf_counter()
#cos2 bias correction
cos2h = cos2_x[:N_curr,i]
cosbcc, cosbc , error = bias_correct_cos2(cos2h,cos2h,eb,norm2dmu_x[i],T,Nsub,M,eta,q,w)
cos2_bc[:N_curr,i] = cosbcc
# print(f"Cos2 bc for N={N_curr} done in {(time.perf_counter()-smalltic2):.2f}s")
with open('status_bc_morevals_allrec.txt', 'a') as f:
f.write(f"\nRec={rec_id}: Cos2 bc for N={N_curr} done in {(time.perf_counter()-smalltic2):.2f}s")
f.write(f"\nTotal time elapsed: {(time.perf_counter()-biggesttic):.2f}s")
np.savez(f'./results_janelia_trialeffect/bc_morevals_eig_cos2_rec{rec_id}.npz', eigx=eigs_x, \
eigbc=eigs_bc, cos2x=cos2_x, cos2bc=cos2_bc, N_vals=N_vals)
# for i in range(len(inds_plot)):
# ex = eigs_x[:N_vals[inds_plot[i]],i]
# cx = cos2_x[:N_vals[inds_plot[i]],i]
# eb = eigs_bc[:N_vals[inds_plot[i]],i]
# cb = cos2_bc[:N_vals[inds_plot[i]],i]
# coefs_x = np.polyfit(np.log(ex), \
# np.log(cx), deg=1)
# coefs_bc = np.polyfit(np.log(eb), \
# np.log(cb), deg=1)
# gamma_x[i] = coefs_x[0]
# gamma_bc[i] = coefs_bc[0]
# alpha_nv = alpha_from_eigs2(eigs_x, np.array(nvals_plot))
# alpha_bc = alpha_from_eigs2(eigs_bc, np.array(nvals_plot))
# np.savez(f'./results_janelia_trialeffect/bc_morevals_eig_cos2_rec{rec_id}.npz', eigx=eigs_x, \
# eigbc=eigs_bc, cos2x=cos2_x, cos2bc=cos2_bc, N_vals=N_vals, gammax=gamma_x, gammabc=gamma_bc, alphax=alpha_nv, alphabc=alpha_bc)
# %%

bias_correct_morevals_allrec.ipynb at commit ec17015, no license · at the source

Overview

Authors: S Amin Moosavi¹, Sai Sumedh R Hindupur², Hideaki Shimazaki^3,4

ORCID iDs: S Amin Moosavi, Sai Sumedh R Hindupur, Hideaki Shimazaki

Institutions: University of California, Los Angeles (United States); Harvard University (United States); Hokkaido University (Japan); Kyoto University (Japan)

Journal: Science advances, volume 12, issue 39, article eadz9632

Dates: received 20 June 2025; accepted 19 August 2026; published online 25 September 2026; in print September 2026

Type: Research article · Language: English

License: CC BY

Identifiers: DOI 10.1126/sciadv.adz9632 · PMID 42789721 · PMCID PMC13614404 · OpenAlex W7214290456

Open access: gold, a free copy (OpenAlex)

Status: code verified

Categories: mouse (organism), systems (subfield)

Methods: Connectivity, Smoothing, state filtering, decompositions, Single-unit activity, calcium imaging, Spectral & time-frequency

MeSH: Models, Neurological*, Neurons*, Visual Cortex*, Animals, Mice (* major topic)

Topic: Neural dynamics and brain function (Cognitive Neuroscience, Neuroscience), according to OpenAlex

Citations: not cited yet (Europe PMC); 44 references in the paper

Abstract

High-dimensional scale-invariant neural activity is ubiquitous across brain regions and species, but its implications for information coding remain unclear. Here, we ask how stimulus information in the high-dimensional activity of mouse V1 scales with neuron number: Does it saturate due to noise correlations or increase without bound as subpopulations grow? Contrary to previous reports, we find that leading noise components that scale linearly with population size, and thus can limit information, are not sufficiently aligned with the signal to impose a bound. This conclusion follows from two scale-invariant power-law properties of neuronal responses in mouse V1: the noise eigenspectrum and alignment of noise components with the signal. We show that population subsampling links the observed power-law exponents to information boundedness and that information scaling depends on the full eigenspectrum rather than its leading modes. Last, we prove that, under subsampling, information-limiting correlations, if present, are differential correlations. Our findings clarify how information scales in high-dimensional neuronal activity under scale-invariant noise.

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

Repositories

Its files are read in the Code ↔ Paper reader above, with 6 matches between paragraphs and lines of code.

Zenodo 19504768

License: CC-BY-4.0

State: the link answers, verified on 28 September 2026

Evidence: files inventoried

Size: 1 file

Software Heritage: not checked

Found in: “Data, code, and materials availability:”

Not found: README, license file, CITATION.cff, environment file, tests, continuous integration, documentation

Tools: NumPy (36 files), Matplotlib (29 files), scikit-learn (23 files), SciPy (15 files), Numba (1 file)

Availability: 1 check, the latest on 28 September 2026: the link answers (HTTP 200)

28 September 2026: the link answers (HTTP 200)

38 files

code/code_janelia_trialeffect_new/alpha_gamma_est.ipynb, Jupyter, 205 lines
code/code_janelia_trialeffect_new/analyse_janelia_mult_trials_all_recordings.ipynb, Jupyter, 145 lines
code/code_janelia_trialeffect_new/bc_alpha_gamma_est_allrec.ipynb, Jupyter, 179 lines
code/code_janelia_trialeffect_new/bc_figure_main_janelia.ipynb, Jupyter, 362 lines
code/code_janelia_trialeffect_new/bias_correct_morevals_allrec.ipynb, Jupyter, 109 lines
code/code_janelia_trialeffect_new/make_main_figure_2.ipynb, Jupyter, 481 lines
code/code_janelia_trialeffect_new/make_supp_figures.ipynb, Jupyter, 394 lines
code/code_janelia_trialeffect_new/normalized_eigs_bias_correct.ipynb, Jupyter, 163 lines
code/code_simdata_alphapos_gamma2/bias_correct.ipynb, Jupyter, 167 lines
code/code_simdata_alphapos_gamma2/directest_simunb_multrep.ipynb, Jupyter, 212 lines
code/code_simdata_alphapos_gamma2/simdata_figures_bc.ipynb, Jupyter, 419 lines
code/code_simdata_unb_alpha0gamma1/alpha_gamma_est.ipynb, Jupyter, 130 lines
code/code_simdata_unb_alpha0gamma1/directest_simunb_multrep.ipynb, Jupyter, 246 lines
code/code_simdata_unb_alpha0gamma1/make_simunb_supp_fig.ipynb, Jupyter, 221 lines
code/code_simdata_unb_alpha0gamma1/pls_simunb_multrep.ipynb, Jupyter, 273 lines
code/code_simdata_unb_alpha0gamma1/scalinganalysis_simunb_multrep.ipynb, Jupyter, 124 lines
code/code_simdata_unb_gamma1/alpha_gamma_est.ipynb, Jupyter, 152 lines
code/code_simdata_unb_gamma1/directest_simunb_multrep.ipynb, Jupyter, 241 lines
code/code_simdata_unb_gamma1/make_simunb_supp_fig.ipynb, Jupyter, 221 lines
code/code_simdata_unb_gamma1/pls_simunb_multrep.ipynb, Jupyter, 273 lines
code/code_simdata_unb_gamma1/scalinganalysis_simunb_multrep.ipynb, Jupyter, 124 lines
code/functions/__init__.py, Python, 1 line
code/functions/bias_correction_functions.py, Python, 193 lines
code/functions/compute_fi.py, Python, 41 lines
code/functions/data_extract.py, Python, 75 lines
code/functions/make_figures.py, Python, 91 lines
code/functions/pls_analysis.py, Python, 352 lines
code/functions/trial_analysis.py, Python, 337 lines
data_simulated_gamma2_manyrep/functions.py, Python, 63 lines
data_simulated_gamma2_manyrep/gen_simdata.py, Python, 100 lines
data_simulated_gamma2_manyrep/generate_data_simulated.ipynb, Jupyter, 39 lines
data_simulated_unb_alpha0gamma1/functions.py, Python, 63 lines
data_simulated_unb_alpha0gamma1/gen_simdata.py, Python, 100 lines
data_simulated_unb_alpha0gamma1/generate_data_simulated.ipynb, Jupyter, 39 lines
data_simulated_unb_gamma1/functions.py, Python, 63 lines
data_simulated_unb_gamma1/gen_simdata.py, Python, 100 lines
data_simulated_unb_gamma1/generate_data_simulated.ipynb, Jupyter, 39 lines
README.md, Text, 4 lines

At the source: doi.org/10.5281/zenodo.19504768

Sai-Sumedh/scaling-population-coding

License: none: the authors keep all their rights

State: the link answers, verified on 28 September 2026

Evidence: files inventoried

Commit: ec170158e146aba4c94296cf2a412d4b4dae8b2f, 24 July 2026

Languages: Jupyter (25), Python (13)

Size: 45 files, 38 scripts

Software Heritage: not archived

Found in: “Data, code, and materials availability:”

Holds: README, 25 notebooks

Not found: license file, CITATION.cff, environment file, tests, continuous integration, documentation

Tools: NumPy (37 files), Matplotlib (29 files), scikit-learn (23 files), SciPy (15 files), Numba (1 file)

Availability: 1 check, the latest on 28 September 2026: the link answers

28 September 2026: the link answers

39 files

code/code_janelia_trialeffect_new/alpha_gamma_est.ipynb, Jupyter, 205 lines
code/code_janelia_trialeffect_new/analyse_janelia_mult_trials_all_recordings.ipynb, Jupyter, 145 lines
code/code_janelia_trialeffect_new/bc_alpha_gamma_est_allrec.ipynb, Jupyter, 179 lines
code/code_janelia_trialeffect_new/bc_figure_main_janelia.ipynb, Jupyter, 362 lines
code/code_janelia_trialeffect_new/bias_correct_morevals_allrec.ipynb, Jupyter, 109 lines, 2 matches
code/code_janelia_trialeffect_new/make_main_figure_2.ipynb, Jupyter, 481 lines
code/code_janelia_trialeffect_new/make_supp_figures.ipynb, Jupyter, 394 lines
code/code_janelia_trialeffect_new/normalized_eigs_bias_correct.ipynb, Jupyter, 163 lines
code/code_janelia_trialeffect_new/num_neurons.ipynb, Jupyter, 25 lines
code/code_simdata_alphapos_gamma2/bias_correct.ipynb, Jupyter, 167 lines, 2 matches
code/code_simdata_alphapos_gamma2/directest_simunb_multrep.ipynb, Jupyter, 212 lines
code/code_simdata_alphapos_gamma2/simdata_figures_bc.ipynb, Jupyter, 419 lines
code/code_simdata_unb_alpha0gamma1/alpha_gamma_est.ipynb, Jupyter, 130 lines
code/code_simdata_unb_alpha0gamma1/directest_simunb_multrep.ipynb, Jupyter, 246 lines
code/code_simdata_unb_alpha0gamma1/make_simunb_supp_fig.ipynb, Jupyter, 221 lines
code/code_simdata_unb_alpha0gamma1/pls_simunb_multrep.ipynb, Jupyter, 273 lines
code/code_simdata_unb_alpha0gamma1/scalinganalysis_simunb_multrep.ipynb, Jupyter, 124 lines
code/code_simdata_unb_gamma1/alpha_gamma_est.ipynb, Jupyter, 152 lines
code/code_simdata_unb_gamma1/directest_simunb_multrep.ipynb, Jupyter, 241 lines
code/code_simdata_unb_gamma1/make_simunb_supp_fig.ipynb, Jupyter, 221 lines
code/code_simdata_unb_gamma1/pls_simunb_multrep.ipynb, Jupyter, 273 lines
code/code_simdata_unb_gamma1/scalinganalysis_simunb_multrep.ipynb, Jupyter, 124 lines
code/functions/__init__.py, Python, 1 line
code/functions/bias_correction_functions.py, Python, 193 lines
code/functions/compute_fi.py, Python, 41 lines
code/functions/data_extract.py, Python, 75 lines, 1 match
code/functions/make_figures.py, Python, 91 lines
code/functions/pls_analysis.py, Python, 352 lines
code/functions/trial_analysis.py, Python, 337 lines
data_simulated_gamma2_manyrep/functions.py, Python, 63 lines
data_simulated_gamma2_manyrep/gen_simdata.py, Python, 100 lines, 1 match
data_simulated_gamma2_manyrep/generate_data_simulated.ipynb, Jupyter, 39 lines
data_simulated_unb_alpha0gamma1/functions.py, Python, 63 lines
data_simulated_unb_alpha0gamma1/gen_simdata.py, Python, 100 lines
data_simulated_unb_alpha0gamma1/generate_data_simulated.ipynb, Jupyter, 39 lines
data_simulated_unb_gamma1/functions.py, Python, 63 lines
data_simulated_unb_gamma1/gen_simdata.py, Python, 100 lines
data_simulated_unb_gamma1/generate_data_simulated.ipynb, Jupyter, 39 lines
README.md, Text, 10 lines

At the source: github.com/Sai-Sumedh/scaling-population-coding

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:

2 repositories of the authors' code, each at its verified commit, with its license and how the link was found in the paper;
75 scripts, each with its path and the digest of its content;
6 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, code, and materials availability

All data and code needed to evaluate and reproduce the results in the paper are present in the paper and/or the Supplementary Materials. The two-photon calcium imaging recordings analyzed in this study are publicly available from Pachitariu et al. (44), “Recordings of 20,000 neurons from V1 in response to oriented stimuli,” https://doi.org/10.25378/janelia.8279387. The code for the data analysis is archived on Zenodo (https://doi.org/10.5281/zenodo.19504768) and linked to the GitHub repository: https://github.com/Sai-Sumedh/scaling-population-coding. This study did not generate new materials.

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, 28 September 2026: the first record

Recorded: type, language, journal, volume, issue, pages, dates, 3 authors, 5 MeSH terms, 42 references.

Cite

This paper

Moosavi, S. A., Hindupur, S. S. R., & Shimazaki, H. (2026). Population coding under the scale invariance of high-dimensional noise. Science advances, 12(39), eadz9632. https://doi.org/10.1126/sciadv.adz9632

BibTeX

@article{moosavi2026population, author = {Moosavi, S Amin and Hindupur, Sai Sumedh R and Shimazaki, Hideaki}, title = {{Population coding under the scale invariance of high-dimensional noise}}, journal = {Science advances}, year = {2026}, month = sep, volume = {12}, number = {39}, pages = {eadz9632}, publisher = {American Association for the Advancement of Science}, issn = {2375-2548}, doi = {10.1126/sciadv.adz9632}, url = {https://doi.org/10.1126/sciadv.adz9632}, pmid = {42789721}, pmcid = {PMC13614404} }

RIS

TY - JOUR AU - Moosavi, S Amin AU - Hindupur, Sai Sumedh R AU - Shimazaki, Hideaki TI - Population coding under the scale invariance of high-dimensional noise T2 - Science advances J2 - Sci Adv PY - 2026 DA - 2026/09/25 VL - 12 IS - 39 SP - eadz9632 SN - 2375-2548 PB - American Association for the Advancement of Science DO - 10.1126/sciadv.adz9632 UR - https://doi.org/10.1126/sciadv.adz9632 LA - en ER -

CSL-JSON

{ "id": "10.1126/sciadv.adz9632", "type": "article-journal", "title": "Population coding under the scale invariance of high-dimensional noise", "container-title": "Science advances", "author": [ { "family": "Moosavi", "given": "S Amin" }, { "family": "Hindupur", "given": "Sai Sumedh R" }, { "family": "Shimazaki", "given": "Hideaki" } ], "container-title-short": "Sci Adv", "volume": "12", "issue": "39", "page": "eadz9632", "DOI": "10.1126/sciadv.adz9632", "PMID": "42789721", "PMCID": "PMC13614404", "ISSN": "2375-2548", "publisher": "American Association for the Advancement of Science", "URL": "https://doi.org/10.1126/sciadv.adz9632", "language": "en", "issued": { "date-parts": [ [ 2026, 9, 25 ] ] } }

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

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.