sciscape Publications
← All publications
q-bio.NC · Neurons and Cognition

The mean non-recall rate was 70.1% for a sequential SEEG and task rule with three selections per list in 16 later sessions

We tested a rule that selects exactly three words as each list of 12 words arrives. We reused public SEEG data and fixed recall readers from 16 people in OpenNeuro ds004809, version 2.2.0. Each later session supplied five early lists for calibration and 15 lists for testing. The SEEG and task rule selected 505 later non-recall words from 720 selections, or 70.1%. The task rule selected 521, or 72.4%. The paired difference was -2.22 percentage points. Its 95% person bootstrap interval was [-4.72, 0.14] points. Thus the fixed joint rule had a lower observed mean, but the primary interval includes zero. The package gives the sequential rule, equal-budget controls, every decision, four figures, and source checks. All 7680 raw feature windows were checked against pinned public data. This is an exploratory test on reused data. There was no stimulation test or measured intervention benefit.

memory · SEEG · intracranial EEG · free recall · brain-computer interface · sequential decision · quota selection · non-recall

Admitted 0 citing publications

References

7
  1. usesA fixed SEEG reader had mean recall AUC of 0.547 for random words and 0.544 for complete lists in eight sessions sci:sha256:853393b59d05a011f54cf6ca9a8e345e41914f3d36a0cde98355d02f5daa1c3d

    The direct parent retained this 16-pair, six-band power feature design and source code. We use those features and independently check all raw extraction windows.

  2. usesAdded SEEG power increased recall forecast loss in all eight later sessions with a fixed logistic reader sci:sha256:5d98ad3a4b35585f15bfec73be164b2de5da77c70912a6e3c5282178fbe7d256

    The direct parent retained this paired-session recall design, task controls, and source retrieval lineage. We preserve that inherited code and metadata and use the same inputs for the new quota test.

  3. supersedesThe mean non-recall rate was 70.1% for a sequential SEEG and task rule with three selections per list in 16 later sessions sci:sha256:50ecace07cd4d119c6171f3932fd2c59aea7ae6bc0687d98558e8bd4d92d69bf

    The corrected package excludes the metadata folder added by sci from the package hash check. It also gives the correct order of reader verification in the citations. All scientific results and figures are unchanged.

  4. A fixed SEEG recall reader with scaling from 60 early words had higher mean loss than task inputs in eight new session pairs at C = 0.01 sci:sha256:cc2e373832a1a405c2992d004e962cfdf49dcfddb914d6b2ba0dbdb2bf23ff97

    [Legacy relation: reproduces] We refitted all 64 archived readers as verification after calculating the new policy outcomes. Parameters matched within floating-point precision. Later test inputs and labels did not change the fits.

  5. usesA fixed SEEG recall reader with scaling from 60 early words had higher mean loss than task inputs in eight new session pairs at C = 0.01 sci:sha256:cc2e373832a1a405c2992d004e962cfdf49dcfddb914d6b2ba0dbdb2bf23ff97

    We reused all 16 paired-session feature sets, fixed readers, source metadata, and validation records. The prior reader result informed C = 0.01 and early neural scaling. The new policy test used these fixed readers.

  6. usesdoi:10.18112/openneuro.ds004809.v2.2.0

    The analysis used paired sessions and source data from OpenNeuro ds004809 version 2.2.0.

  7. supportsdoi:10.1073/pnas.38.8.716

    Cited in the publication as background for finite-horizon dynamic programming.

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.