Atlas health

Operational status of the atlas: pooled-vs-by-subject reconciliation, open curation tasks, external data blockers, the model-coverage roadmap, and internal link integrity. CI runs each check on every commit; this page is the latest committed result.

The shape of the atlas under check

  1. Family 7 Reusable operational task idea.
  2. Protocol 18 One concrete variant with species, stimulus, choice.
  3. Dataset 16 Pointer to public or private behavioural data.
  4. Slice 17 One protocol × dataset bound to a reproducible pipeline.
  5. Finding 91 One curve, on shared signed-evidence axes.

Where do we stand?

Reproducibility audit is ok; the curation queue has 6 open tasks; 1 blocker group are waiting on external data (8 high-priority); the model-coverage roadmap has 21 items; link integrity finds 0 issues.

Audit

ok Curation open 6 Blocker groups 1 Roadmap items 21 Link issues 0

What's blocked, what's ready?

One board for local implementation targets, external data blockers, and extraction gaps. Rows are generated from the curation queue, data-request records, and model-coverage roadmap.

Coverage Gap Matrix

Visual triage across source data, trial tables, extracted findings, model fits, reports, and request state.

13/13
ok part ready miss block n/a
  • Busse 2011 mouse visual contrast port-task trial data

    paper.busse-2011-detection-visual-contrast

    blocked
    Raw
    block
    Trials
    block
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    miss
    Request
    ready

    author request raw trials

    Send the ready-to-send request draft and record a sent event.

  • Khalvati-Kiani-Rao raw behavioral MATLAB files

    paper.khalvati-kiani-rao-2021

    blocked
    Raw
    part
    Trials
    block
    Figure
    ok
    Findings
    part
    Models
    part
    Reports
    ok
    Request
    block

    author request raw trials

    Resolve the recorded blocker before continuing the request.

  • Lak 2020 mouse visual contrast wheel behavior and value-block data

    paper.lak-2020-reinforcement-biases-confidence

    blocked
    Raw
    block
    Trials
    block
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    miss
    Request
    ready

    author request raw trials

    Send the ready-to-send request draft and record a sent event.

  • Burgess 2017 high-yield visual psychophysics source data

    paper.burgess-2017-high-yield-visual-psychophysics

    blocked
    Raw
    block
    Trials
    block
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    miss
    Request
    part

    author request source data

    Complete the request draft before sending.

  • Pho 2018 visual contrast go/no-go data and MATLAB code

    paper.pho-2018-task-dependent-parietal

    blocked
    Raw
    block
    Trials
    block
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    miss
    Request
    ready

    author request source data

    Send the ready-to-send request draft and record a sent event.

  • Mouse port-based visual contrast 2AFC task

    protocol.mouse-visual-contrast-port-2afc

    extraction
    Raw
    miss
    Trials
    miss
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    unk
    Request
    ready

    needs dataset

    Find or curate an open dataset for this protocol, then add reciprocal protocol/dataset metadata.

  • Mouse port-based visual contrast 2AFC task

    protocol.mouse-visual-contrast-port-2afc

    extraction
    Raw
    unk
    Trials
    miss
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    miss
    Request
    ready

    needs vertical slice

    Choose a dataset-backed instance and add a vertical slice with analysis artifacts.

  • Mouse visual contrast lick go/no-go task

    protocol.mouse-visual-contrast-lick-gonogo

    extraction
    Raw
    miss
    Trials
    miss
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    unk
    Request
    ready

    needs dataset

    Find or curate an open dataset for this protocol, then add reciprocal protocol/dataset metadata.

  • Mouse visual contrast lick go/no-go task

    protocol.mouse-visual-contrast-lick-gonogo

    extraction
    Raw
    unk
    Trials
    miss
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    miss
    Request
    ready

    needs vertical slice

    Choose a dataset-backed instance and add a vertical slice with analysis artifacts.

  • Mouse visual contrast wheel task with reward-value blocks

    protocol.mouse-visual-contrast-wheel-value-blocks

    extraction
    Raw
    miss
    Trials
    miss
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    unk
    Request
    ready

    needs dataset

    Find or curate an open dataset for this protocol, then add reciprocal protocol/dataset metadata.

  • Mouse visual contrast wheel task with reward-value blocks

    protocol.mouse-visual-contrast-wheel-value-blocks

    extraction
    Raw
    unk
    Trials
    miss
    Figure
    unk
    Findings
    miss
    Models
    miss
    Reports
    miss
    Request
    ready

    needs vertical slice

    Choose a dataset-backed instance and add a vertical slice with analysis artifacts.

  • Allen Visual Behavior Neuropixels Change Detection

    slice.allen-visual-behavior-neuropixels

    ready
    Raw
    ok
    Trials
    ok
    Figure
    n/a
    Findings
    miss
    Models
    ready
    Reports
    ok
    Request
    n/a

    missing stimulus value

    Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice.

  • Rodgers Whisker Object Recognition

    slice.rodgers-whisker-object-recognition

    ready
    Raw
    ok
    Trials
    ok
    Figure
    n/a
    Findings
    miss
    Models
    ready
    Reports
    ok
    Request
    n/a

    missing stimulus value

    Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice.

Are the curves reproducible?

For each (paper × curve type × condition) group with both a pooled finding and ≥2 per-subject findings, the audit recomputes the n-weighted aggregate of the subject curves and compares to the pooled curve at every matching x.

ok

5 groups audited.

max |diff| = 0.000000 (tolerance 0.0100)

local: uv run behavtaskatlas audit-findings

Paper Curve Condition Subjects x overlap max |diff| Status
paper.palmer-huk-shadlen-2005 psychometric — 6 11 0.000000 ok
paper.roitman-shadlen-2002 psychometric — 2 11 0.000000 ok
paper.walsh-2024-prior-cue psychometric cue=invalid 12 38 0.000000 ok
paper.walsh-2024-prior-cue psychometric cue=neutral 12 38 0.000000 ok
paper.walsh-2024-prior-cue psychometric cue=valid 12 38 0.000000 ok

What needs cleaning?

6 open of 6 total · generated 2026-07-26T14:22:00.698814+00:00. Items come from the relationship-graph QA pass: orphan records, missing reciprocal links, coverage gaps.

By priority

  • normal 6

By action type

  • needs dataset 3
  • needs vertical slice 3

normal priority (6)

  • needs dataset normal

    Protocol has no linked dataset record yet.

    Next: Find or curate an open dataset for this protocol, then add reciprocal protocol/dataset metadata.

    protocol.mouse-visual-contrast-port-2afc

  • needs dataset normal

    Protocol has no linked dataset record yet.

    Next: Find or curate an open dataset for this protocol, then add reciprocal protocol/dataset metadata.

    protocol.mouse-visual-contrast-lick-gonogo

  • needs dataset normal

    Protocol has no linked dataset record yet.

    Next: Find or curate an open dataset for this protocol, then add reciprocal protocol/dataset metadata.

    protocol.mouse-visual-contrast-wheel-value-blocks

  • needs vertical slice normal

    Protocol has no report-backed vertical slice yet.

    Next: Choose a dataset-backed instance and add a vertical slice with analysis artifacts.

    protocol.mouse-visual-contrast-port-2afc

  • needs vertical slice normal

    Protocol has no report-backed vertical slice yet.

    Next: Choose a dataset-backed instance and add a vertical slice with analysis artifacts.

    protocol.mouse-visual-contrast-lick-gonogo

  • needs vertical slice normal

    Protocol has no report-backed vertical slice yet.

    Next: Choose a dataset-backed instance and add a vertical slice with analysis artifacts.

    protocol.mouse-visual-contrast-wheel-value-blocks

Which data can't we publish?

data_requests.csv

Roadmap rows that cannot be closed by another local fit or adapter pass — the missing source files have to be requested or located. Tracked separately so proxy-backed model wins remain visible.

Blocker groups 1
Roadmap rows 8 high priority 8
Findings affected 8
Active requests 5 ready 3
Follow-up due 0

Request queue

Queue state

ready to send

3
  • Busse 2011 mouse visual contrast port-task trial data

    busse-2011-visual-contrast-port-trials

    high

    Send the ready-to-send request draft and record a sent event.

    Last event
    drafted · 87 day(s)
    Follow-up
    none 
    uv run behavtaskatlas data-request-event data_request.busse-2011-visual-contrast-port-trials --event-type sent --event-date 2026-07-26 --actor "curator" --notes "Sent the author request." --status requested --next-follow-up-date 2026-08-09 --create-evidence-stub
  • high

    Send the ready-to-send request draft and record a sent event.

    Last event
    drafted · 87 day(s)
    Follow-up
    none 
    uv run behavtaskatlas data-request-event data_request.lak-2020-visual-contrast-wheel-source-data --event-type sent --event-date 2026-07-26 --actor "curator" --notes "Sent the author request." --status requested --next-follow-up-date 2026-08-09 --create-evidence-stub
  • Pho 2018 visual contrast go/no-go data and MATLAB code

    pho-2018-visual-contrast-gonogo-source-data

    normal

    Send the ready-to-send request draft and record a sent event.

    Last event
    drafted · 87 day(s)
    Follow-up
    none 
    uv run behavtaskatlas data-request-event data_request.pho-2018-visual-contrast-gonogo-source-data --event-type sent --event-date 2026-07-26 --actor "curator" --notes "Sent the author request." --status requested --next-follow-up-date 2026-08-09 --create-evidence-stub

Queue state

blocked

1

Queue state

draft

1

author request raw trials

Khalvati-Kiani-Rao macaque RDM confidence source data

non-human-primate · Macaque random-dot motion confidence wagering task · figure-source-data

blocked external data

Blocker

The Nature source-data ZIP and POMDP-Confidence code archive provide figure source tables and code, but not the behavioral MATLAB files referenced by the code (`beh_data.monkey1.mat` and `beh_data.monkey2.mat`). The article states that analyzed data are available from R.K. on reasonable request.

Next action

Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

Khalvati K, Kiani R, Rao RPN macaque-rdm-confidence-wager source package

Tracked request

Khalvati-Kiani-Rao raw behavioral MATLAB files

request .md blocked

Replace figure-source and direction-choice proxy fits with harmonized trial-level records for the Kiani-Shadlen macaque RDM confidence task, so the atlas can fit psychometric and DDM models against real choice/RT outcomes rather than inferred accuracy summaries.

Resolve the recorded blocker before continuing the request.

Timeline (2)
  1. drafted 2026-04-29 behavtaskatlas

    Drafted from the generated model-roadmap blocker after verifying that the public source-data ZIP and POMDP-Confidence code archive do not include the behavioral MATLAB files referenced by the code.

    data_requests/khalvati-kiani-rao-2021-raw-behavior-matlab.yaml

  2. note 2026-04-29 behavtaskatlas

    Paused the outbound author request while the roadmap shifts to broader model-selection comparability work. Keep the draft available for a later Khalvati raw-trial recovery pass.

Ready-to-send request draft

Request for Khalvati/Kiani/Rao 2021 macaque behavioral MATLAB files 

Dear Dr. Kiani / Khalvati / Rao team,

I am curating behavtaskatlas, an open-science atlas of sensory-guided decision-making tasks and reusable behavioral-model fits.

I am working with the public source-data ZIP and POMDP-Confidence code for:

Khalvati K, Kiani R, Rao RPN. Bayesian inference with incomplete knowledge explains perceptual confidence and its deviations from accuracy. Nature Communications, 2021.

The public code appears to reference behavioral MATLAB files named `beh_data.monkey1.mat` and `beh_data.monkey2.mat`, but I could not find those files in the Nature source-data ZIP or the public POMDP-Confidence code archive. Would you be willing to share those analyzed/raw behavioral MATLAB files, or point me to their current public location?

The immediate use is to build a provenance-preserving trial-level harmonizer for the macaque RDM confidence task, recover signed choice and response-time fields where available, and fit psychometric / DDM models with explicit citation and caveats. I would also appreciate any redistribution terms you want observed for derived canonical trial tables or aggregate model-fit outputs.

Thank you,
behavtaskatlas curator
Rank Finding Curve Scope Impact
1 khalvati-kiani-rao-2021.accuracy.accuracy-no-sure-target.m1 accuracy by strength accuracy summary Improves cross-paper model comparisons by reducing source and proxy caveats.
2 khalvati-kiani-rao-2021.accuracy.accuracy-no-sure-target.m2 accuracy by strength accuracy summary Improves cross-paper model comparisons by reducing source and proxy caveats.
3 khalvati-kiani-rao-2021.accuracy.accuracy-sure-available-direction-chosen.m1 accuracy by strength accuracy summary Improves cross-paper model comparisons by reducing source and proxy caveats.
4 khalvati-kiani-rao-2021.accuracy.accuracy-sure-available-direction-chosen.m2 accuracy by strength accuracy summary Improves cross-paper model comparisons by reducing source and proxy caveats.
5 khalvati-kiani-rao-2021.chronometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m1 chronometric chronometric summary Improves cross-paper model comparisons by reducing source and proxy caveats.
6 khalvati-kiani-rao-2021.chronometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m2 chronometric chronometric summary Improves cross-paper model comparisons by reducing source and proxy caveats.
7 khalvati-kiani-rao-2021.psychometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m1 psychometric direct choice Improves cross-paper model comparisons by reducing source and proxy caveats.
8 khalvati-kiani-rao-2021.psychometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m2 psychometric direct choice Improves cross-paper model comparisons by reducing source and proxy caveats.

What's the next pass?

Ranked from coverage gaps, comparison-scope warnings, caveats, and near-miss slice capabilities. Top 20 shown; full export: model_roadmap.csv.

high 8 medium 0 low 13 blocked 8
Rank Target Issue Action Impact

1

high blocked external data 		khalvati-kiani-rao-2021.accuracy.accuracy-no-sure-target.m1

finding

 proxy/source data Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

author request raw trials

Improves cross-paper model comparisons by reducing source and proxy caveats.

2

high blocked external data 		khalvati-kiani-rao-2021.accuracy.accuracy-no-sure-target.m2

finding

 proxy/source data Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

author request raw trials

Improves cross-paper model comparisons by reducing source and proxy caveats.

3

high blocked external data 		khalvati-kiani-rao-2021.accuracy.accuracy-sure-available-direction-chosen.m1

finding

 proxy/source data Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

author request raw trials

Improves cross-paper model comparisons by reducing source and proxy caveats.

4

high blocked external data 		khalvati-kiani-rao-2021.accuracy.accuracy-sure-available-direction-chosen.m2

finding

 proxy/source data Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

author request raw trials

Improves cross-paper model comparisons by reducing source and proxy caveats.

5

high blocked external data 		khalvati-kiani-rao-2021.chronometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m1

finding

 proxy/source data Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

author request raw trials

Improves cross-paper model comparisons by reducing source and proxy caveats.

6

high blocked external data 		khalvati-kiani-rao-2021.chronometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m2

finding

 proxy/source data Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

author request raw trials

Improves cross-paper model comparisons by reducing source and proxy caveats.

7

high blocked external data 		khalvati-kiani-rao-2021.psychometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m1

finding

 proxy/source data Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

author request raw trials

Improves cross-paper model comparisons by reducing source and proxy caveats.

8

high blocked external data 		khalvati-kiani-rao-2021.psychometric.direction-choice-proxy.accuracy-no-sure-target.kiani-shadlen-m2

finding

 proxy/source data Request the raw behavioral MATLAB files from the Kiani lab/R.K.; keep figure-source and proxy caveats prominent until those files can be harmonized.

author request raw trials

Improves cross-paper model comparisons by reducing source and proxy caveats.

9

low ready 		allen-visual-behavior-neuropixels

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

10

low ready 		allen-visual-behavior-neuropixels

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

11

low ready 		allen-visual-behavior-neuropixels

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

12

low ready 		allen-visual-behavior-neuropixels

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

13

low ready 		rodgers-whisker-object-recognition

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

14

low ready 		rodgers-whisker-object-recognition

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

15

low ready 		rodgers-whisker-object-recognition

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

16

low ready 		rodgers-whisker-object-recognition

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

17

low ready 		rodgers-whisker-object-recognition

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

18

low ready 		rodgers-whisker-object-recognition

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

19

low ready 		rodgers-whisker-object-recognition

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.

20

low ready 		rodgers-whisker-object-recognition

slice

 near-miss slice Add the missing harmonized capability (stimulus_value) or mark the variant as intentionally inapplicable for this slice. Expands the fittable model set for an existing vertical slice.