With shared input history, singular value clipping bounded state sensitivity in 36 fixed 100 ms SEEG readers
With shared input history, a projected norm guard bounded state sensitivity without the next recovery time. It clipped the singular values of each fixed 100 ms endpoint map to 0.95. The projection made the least Frobenius matrix change among all maps with this limit. The test used 36 fixed pairs from nine public SEEG rest records in OpenNeuro ds006065 and ds005670. Six random schedules supplied dense, mixed, and sparse recovery intervals. Only two maps changed. The original readers showed exponential state growth in both pairs under dense gaps; both the scalar and projected guards removed that growth. Mean dense phase error was 68.24 degrees for projection, 68.17 for scalar scaling, and 79.08 for voltage hold. Projection minus scalar scaling was +0.067 degrees. Its descriptive 95 percent record interval was -0.009 to +0.210 degrees. There was no clear phase gain over scalar scaling. Projection added 1.689 degrees against the original reader under sparse gaps. These are reused clinical records with simulated gaps and a numerical filter target. Neural phase truth and stimulation outcomes were not tested. The package preserves voltage, fixed models, code, source receipts, proof, all-seed scores, and three figures. Fresh source checks matched all 7,200,000 voltage values. Independent replay, streaming checks, and exact numerical reruns passed.
state sensitivity · singular value clipping · endpoint norm · SEEG · irregular gaps · phase error · matrix projection
References
8- usesTwo of 36 learned SEEG state maps were unstable during 100 ms simulated gaps every 120 ms sci:sha256:28dec263a684c3edc55154127a63ea5a4656288fb3acc603cfe9f34ad9e44956
We retained all nine voltage extracts, 36 fixed 100 ms pair models, the filter and feature definitions, and source records. We added matrix projection and random recovery intervals with no next-gap time input.
- usesA norm guard bounded state sensitivity for 36 fixed 100 ms SEEG gap readers with common input history sci:sha256:084d1a15d4ea72bf445fccee855ee465a3f0043990104f3832626b39ecb15173
This work established the scalar endpoint norm guard without future recovery timing. We used the same centered scalar rule as a comparator. We tested a full matrix projection and random recovery intervals.
- supportsProducts of two stable cycle maps were unstable in 11 of 36 fitted SEEG readers on a fixed timing grid sci:sha256:d24b68b6fd2c99990305092dbb899ded1d0e9c76ef73b3102b177ba4c068f9b3
Our proof and fixed 100 ms replay checks support the shared-history endpoint norm argument. This study does not test that publication's variable-gap grid. It adds the least matrix change projection and direct accuracy comparison.
- usesA learned state reduced mean 7 Hz filter phase error after 100 ms simulated gaps in nine SEEG rest records with known initial states sci:sha256:101556a678d7a1131cc51640655ad1744d73bf01ea5d5882f68258510d434efd
The fixed 27-feature model weights, calibration, and fitting times came from this isolated-gap study. We did not fit the models again.
- usesA shared reader had 58.38 degrees of mean absolute phase error in two new SEEG rest records at 50 ms with a centered theta reference sci:sha256:6bd2f301da5cc69ca85251358ec5d79565bf3c48c78038666bfdd36352965b56
We retained the corrected EDF channel mapping and native voltage extracts through the parent. Fresh public-prefix checks matched the selected arrays.
- usesTwo exact bounds certified a 7 Hz SEEG filter phase after simulated gaps in nine rest records under voltage caps sci:sha256:1e7436d8e30edf812519ca308906d17bd0b1a757df88c2ccc6c9d70a987f5cb5
The public-prefix reconstruction code and source hash chain came from this work through the parent. We adapted the paths and ran the recipe against fresh public files.
- usesdoi:10.18112/openneuro.ds006065.v1.0.0
The fixed 100 ms models and seven selected records from this version were reused.
- usesdoi:10.18112/openneuro.ds005670.v1.0.0
Two selected records from this version were reused.
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesA constrained fit reduced mean 7 Hz filter phase error against a guard applied after fitting in nine SEEG records with simulated mixed gaps and known initial states
The added test extends the centered singular value projection to four completed gap lengths. It compares that projected guard with the constrained fit.