XDAWNCov+TS+LRClassical method · 4 filters · 50 Hz · primary row · Local Ubuntu / CPU
Supervised fit
61.9%58.8%–65.0%
0.615Mean across held-out participants
19
21
—
xDAWN+LDAClassical method · 4 filters · 50 Hz · secondary row · Local Ubuntu / CPU
Supervised fit
60.4%57.7%–63.2%
0.601Mean across held-out participants
19
21
—
ShallowFBCSPNetCompact model · matrix recipe · AdamW · Local Ubuntu / CUDA
Scratch · 20 epochs
49.7%49.0%–50.4%At or below chance level
0.427Mean across held-out participants
19
21
—
Deep4NetCompact model · matrix recipe · AdamW · Local Ubuntu / CUDA
Scratch · 20 epochs
50.8%50.0%–51.6%
0.409Mean across held-out participants
19
21
—
CSP+LDAClassical method · broadband CSP · shrinkage LDA · Local Ubuntu / CPU
Supervised fit
50.0%49.1%–50.8%At or below chance level
0.453Mean across held-out participants
19
21
—
Scoring time covers fitting and prediction in this configuration, may include accelerator waiting, and excludes data preparation. Scoring time is not recorded for the rows of core-matrix-update-20261106.
Not applicable under this protocol:Standard CCA, FBCCA (they need the stimulation frequencies, and this protocol has none).
Inspect point estimates and 95% intervals side by side on a shared axis. Paired differences appear only where a paired interval is published. Dropdowns are grouped by family alphabetically — never ordered by score.
One fixed configuration per method: library or MOABB-reference settings with the stated deviations. The 4–40 Hz band-pass (or demeaning plus 50 Hz decimation on the ERP protocols) is part of the recipe and is applied to each window on its own.
XDAWNCov+TS+LR: Primary row · one fixed configuration (MOABB-reference settings)
xDAWN+LDA: Secondary row · one fixed configuration (MOABB-reference settings)
ShallowFBCSPNet, Deep4Net and CSP+LDA under the matrix recipe
Deep4Net: Braindecode's automatically scaled Deep4Net for 1-s input (no temporal pooling, 229-sample classifier, 235,752 parameters), not the original Deep4Net configuration. The interval's lower end, 50.04 %, prints as 50.0; it sits 0.04 pp above the 50 % reference.
CSP+LDA: Broadband CSP without band-pass filtering, the recipe of the published mi-rest CSP+LDA cell; not a tuned CSP. On this ERP protocol it reads band power, not the waveform.
Paired differences
Each difference is the mean over people of each person’s difference in balanced accuracy, in percentage points (pp), on the same windows and people.
Paired participant bootstrap (10,000 draws). The four contrasts against the Spectral ridge cells carry 98.75 % intervals (Bonferroni over four); the other contrasts carry nominal 95 % intervals with no familywise claim. 'Higher' or 'lower' means the interval lies wholly above or below zero; 'unresolved' means it includes zero.
ShallowFBCSPNet, Deep4Net and CSP+LDA under the matrix recipe
against the matrix's one-seed EEGNet cell; training variance not included
Difference
Mean (pp)
95% interval, unadjusted
Bonferroni interval over 18 contrasts
Reading
CSP+LDA − EEGNet
−3.3 pp
−5.4 to −1.6
−6.6 to −0.8
lower
Deep4Net − EEGNet
−2.5 pp
−4.2 to −0.9
−5.2 to −0.2
lower
ShallowFBCSPNet − EEGNet
−3.6 pp
−5.4 to −1.9
−6.3 to −0.9
lower
The paired 95 % interval governs the wording; the Bonferroni interval beside it covers all 18 contrasts. Read at 95 % without correction, about one in 20 contrasts with no true difference may show one by chance.
Re-training on the CUDA card is not bit-exact; re-training the published semantic-target EEGNet cell there moved it by 0.1 pp.
Beside the core matrix
Log band-power ridge (5 bands, visible channels)
Five log band powers (1–4, 4–8, 8–13, 13–30 and 30–45 Hz) per channel on all recorded channels of the protocol, from one Hann periodogram of each published window after removing each channel's window mean and a per-window robust amplitude scaling, read by the core ridge head (alpha 100) on the published windows and participant-held-out folds. It uses the same windows, people, folds and ridge head as the Spectral ridge row, repeated from the table above.
Each difference is this row minus the named row on the same people, with a paired participant bootstrap (10,000 draws): an interval wholly above zero reads above, wholly below reads below, and one that includes zero reads not separated, which does not show the two are equivalent. The intervals are descriptive, conditional on these folds, with no multiplicity correction.
Minus the Spectral ridge row: −0.52 pp (−2.42 to +1.43): not separated. The two rows differ in bands, normalisation, estimator, frequency range and sampling rate at once, so the difference measures the whole recipe.
Minus CBraMod, the highest of the 13 frozen-encoder rows this comparison was specified against, chosen on these same test folds: −4.52 pp (−6.23 to −2.78): below. Choosing it on the test folds leans the comparison against this row.
Notes on this comparator
On these 1-second windows the 1–4 Hz band partly carries slow drift within the window and the slow ERP itself, as the Spectral ridge row’s lowest bins do.
Each encoder replaces LaBraM or CBraMod in the frozen recipe above; the windows, people, folds, ridge head and scoring stay the same. Rows are grouped by model family; a row is above or below another only where their 95% intervals do not overlap. Each interval is a descriptive participant bootstrap with no multiplicity correction, and with 126 frozen cells an occasional non-overlap is expected by chance. Every encoder sees this protocol’s published windows, usually shorter than its pretraining context.
Spectral ridge
51.0% (49.6%–52.4%)
Temporal ridge
53.8% (51.9%–55.6%)
LaBraM
49.4% (47.7%–51.2%)
CBraMod
55.0% (53.3%–56.7%)
EEGNet
53.3% (51.7%–54.9%)
REVE Base
50.5% (48.5%–52.3%)
REVE Large
54.6% (52.7%–56.5%)
LUNA Base
48.7% (46.3%–51.1%)
LUNA Large
48.8% (46.9%–50.8%)
CodeBrain
50.8% (49.3%–52.2%)
EEGMamba
52.5% (51.4%–53.5%)
ST-EEGFormer Base
53.6% (52.3%–54.8%)
ST-EEGFormer Large
52.7% (51.3%–54.2%)
eeg-fm-masking MAE (r=9 cm, L=2)
49.5% (48.1%–51.0%)
ERP-FM Base
51.2% (49.3%–53.2%)
SingLEM
51.2% (49.2%–53.1%)
ZUNA 1.1
51.6% (50.3%–53.0%)
0%25%50%75%100%
Balanced accuracy under this protocol. Grey: the released rows above; rust: the further encoders, in family order. Dot: the estimate; line: 95% interval; dashed line: chance level.
Not runBrainTokenizer needs a 512-sample window (2.0 s at 256 Hz); the published segment is 1 s = 256 samples, so upstream unfold() would zero-pad 256 of 512 samples of every window (invented samples). All upstream downstream tasks use >= 2 s windows.
19
—
not in the authors' published pretraining list (checked 2026-10-04)source ↗
not in the authors' published pretraining list (checked 2026-10-04)source ↗
ZUNA 1.113Foundation modelModel card: research use only, not for diagnosis or clinical use.
Frozen encoder + ridge head
51.6%50.3%–53.0%
0.511
19
21
Exposure unknownunknown: the authors do not list their pretraining data closely enough to decidesource ↗
Pretraining exposure is what each model’s authors publish as their pretraining list, checked on 2026-10-04: a dataset absent from it is “not in the authors’ published pretraining list”, which is not proof that its recordings were never seen. ST-EEGFormer lists BETA and SingLEM lists TMNRED; ZUNA 1.1 publishes no list, so every one of its cells is unknown; EEGMamba’s list is read from its official code (its paper was not read), with medium confidence.
Row footnotes
Input removes the per-segment per-channel mean before the published /100 scaling (declared before scoring); a no-mean-removal sensitivity run scores P300 0.71 percentage points higher and sleep 1.40 lower.
Input removes the per-segment per-channel mean (declared before scoring); the no-mean-removal sensitivity run scores P300 2.06 percentage points lower and sleep 0.93 lower.
Resampled to 256 Hz and truncated to whole 40-sample patches; 1-2 s windows and 4-8 channel montages are outside the authors' tested regime.
As LUNA Base. Frozen probes only.
p300-target and semantic-target not run: the 2 s tokenizer window cannot be met by 1 s segments without zero padding. Upstream common average reference kept on 4-8 channel montages.
Published CBraMod pooling (patch mean, channels kept); LoRA effectively adapts only the V projection (1-token attention window upstream).
Published CBraMod pooling; LoRA on Mamba2 in_proj and out_proj (24 targets).
BETA cells are pretraining-exposed (BETA is in the ST-EEGFormer corpus).
BETA cells are pretraining-exposed. The authors' downstream class (1-based temporal index) was used; Large may have been pretrained with a 0-based index. Frozen probes only.
Paper-recommended masking geometry; channel-retained pooling (C x 512). Three sibling checkpoints appear in the masking ablation panel.
In design: p300-target and semantic-target. The other six protocols are negative controls outside the model's design. ERP windows have no pre-stimulus baseline.
semantic-target is pretraining-exposed (TMNRED is in the SingLEM corpus). Each channel is encoded independently; the last 0.25 s of each 2 s segment is not seen by the tokenizer.
Pretraining exposure unknown for every core dataset (no published source list). Encoder only; no channel upsampling or reconstruction.
A sibling of the paper-recommended checkpoint in the masking ablation: the same architecture, corpus and recipe, another masking framework or geometry.
Masking ablation: three sibling checkpoints of eeg-fm-masking
Intervals are marginal: the same people under two checkpoints are not tested as a pair.
not in the authors' published pretraining list (checked 2026-10-04)source ↗
Weights licences and terms
REVE Base · REVE Large: REVE Responsible Use License v1.0 (owner-approved 2026-10-04)REVE Responsible Use License v1.0: aggregate scientific results only; no re-identification, model inversion, membership inference or non-consensual profiling; cite the REVE paper and name the model version; no adapted weights shared.Model versions: REVE Base @ dc2a075c · REVE Large @ 317531c7.paper ↗
LUNA Base · LUNA Large: CC BY-ND 4.0 (owner-approved: internal adaptation, aggregate scores only, no adapted weights or LoRA deltas shared)CC BY-ND 4.0 weights: scores from internal probing/adaptation only; no modified weights, LoRA or adapter deltas are shared; no implied endorsement by the LUNA authors; the LUNA name is not used to suggest an official release.paper ↗
ST-EEGFormer Base · ST-EEGFormer Large: MIT (repo LICENSE; README reserves the paper, diagrams and the name)MIT (repository LICENSE; the README reserves the paper, diagrams and the ST-EEGFormer name).paper ↗
eeg-fm-masking MAE (r=9 cm, L=2) · eeg-fm-masking JEPA (r=9 cm, L=2) · eeg-fm-masking MAE (r=one channel, L=1) · eeg-fm-masking JEPA (r=one channel, L=1): CC-BY-4.0 (cite arXiv:2609.33487)CC-BY-4.0 weights: attribution required (cite arXiv:2609.33487).paper ↗
SingLEM: MIT (checkpoint committed in the MIT repository)MIT.paper ↗
ZUNA 1.1: Apache-2.0; model card: research use only, not for diagnosis or clinical useApache-2.0 weights; model card states research use only, not for diagnosis or clinical use.paper ↗
No model’s authors endorse these results, and no adapted weights or LoRA deltas were kept or shared.
Notes on these rows
REVE Base: REVE input removes the per-segment per-channel mean (declared before scoring; the published LaBraM/CBraMod rows did not); a no-mean-removal sensitivity run is stated in the row footnote.
REVE Large: REVE input removes the per-segment per-channel mean (declared before scoring); a no-mean-removal sensitivity run is stated in the row footnote.
LUNA Base: LUNA resamples to 256 Hz and keeps whole 40-sample patches: 1 s keeps 240/256 samples, 2 s keeps 480/512. Point estimate below chance; reported as is.
LUNA Large: LUNA resamples to 256 Hz and keeps whole 40-sample patches: 1 s keeps 240/256 samples, 2 s keeps 480/512. Point estimate below chance; reported as is.
ST-EEGFormer Large: ST-EEGFormer Large uses the authors' downstream class (1-based temporal index); a label-free reconstruction check suggests Large was pretrained with a 0-based index, so its temporal sinusoid may be shifted by one position.
ERP-FM Base: In ERP-FM design (post-onset 1 s windows without baseline; pretraining used baselined, filtered epochs).
SingLEM: SingLEM tokenizer (128-sample tokens, stride 96 at 128 Hz) leaves the last 0.25 s of every 2 s segment unseen.
ZUNA 1.1: ZUNA pretraining sources are not published: exposure unknown on every core dataset. 0-4 electrode coordinates per protocol are clamped to the edge of the +-0.12 m position grid as upstream does.
EEGPT ran in the same runs as the published LaBraM and CBraMod rows above, with only the encoder swapped: the same windows, people, folds, ridge head and scoring.
Spectral ridge
51.0% (49.6%–52.4%)
Temporal ridge
53.8% (51.9%–55.6%)
LaBraM
49.4% (47.7%–51.2%)
CBraMod
55.0% (53.3%–56.7%)
EEGNet
53.3% (51.7%–54.9%)
EEGPT
54.0% (51.5%–56.6%)
0%25%50%75%100%
Balanced accuracy under this protocol. Grey: the released rows above; rust: EEGPT. Dot: the estimate; line: 95% interval; dashed line: chance level.
not in the authors' published pretraining list (checked 2026-10-12)source ↗
Footnotes to this row
On the P300 and semantic targets, one-second windows ran as four real 0.25-second patches: the patch grid was set to four after the encoder was built, with no padding. EEGPT’s encoder treats each time patch on its own, so this is numerically the same as building it for 256-sample inputs, as the Orthosis pilot built it for its 128-sample windows.
Every core window is shorter than EEGPT’s 4-second pretraining window: 1 or 2 seconds, and sleep as fifteen 2-second segments. The encoder computes each 0.25-second patch the same way at any length; a shorter window gives fewer patches to average.
The readout is the site’s fixed probe on the frozen summary tokens, as for the published LaBraM and CBraMod rows: ridge (alpha 100) on 2,048 pooled values, and on idle a per-person logistic head. It is not the authors’ downstream probe, a learnable spatial filter with a linear head, and the windows are fed without the average reference EEGPT was pretrained with.
Against the published rows
Each row of the table above that experiments.json holds, by whether its 95% interval and EEGPT’s overlap (no multiplicity correction):
Spectral ridge: the intervals overlap
Temporal ridge: the intervals overlap
LaBraM: EEGPT above, the intervals 0.28 pp apart
CBraMod: the intervals overlap
EEGNet: the intervals overlap
EEGPT’s exposure is read from its authors’ published pretraining list, checked on 2026-10-12, as for the further encoders. Its pretraining corpus: Five public datasets, as the paper’s Table 1 lists them: PhysioMI (PhysioNet motor imagery and execution), HGD, TSU (the Tsinghua 40-target SSVEP benchmark), SEED and M3CV; 58 electrodes, 256 Hz, millivolts, average reference, 4-second windows.
Licence and terms
Code licence: Apache-2.0 (the repository’s source code only)
Weights licence: unresolved: the checkpoint’s own licence and terms could not be read from its hosting record, and the repository’s code licence is not evidence that it covers the weights
Weights: no original or converted weights are distributed; the hashes identify the checkpoint used and grant nothing
Rights review: Published by the owner’s decision, aggregate results only, as the pilot comparison is. The checkpoint’s licence could not be read; no publisher prohibition on independently computed aggregate results was found; no original or converted weights are distributed.
Six LaBraM at Scale encoders and REVE Small’s encoder in the same frozen recipe as the rows above: the same windows, people, folds, head and scoring, with no random-initialization control.
Balanced accuracy with its 95% interval, and macro F1, frozen encoders.
its pretraining corpus includes this protocol's dataset
Molecule-LLocal Ubuntu / CPU
51.4%49.2%–53.5%interval includes chance
0.509
its pretraining corpus includes this protocol's dataset
REVE Small · early checkpoint (epoch 15 of 200)
REVE SmallLocal Ubuntu / CPU
54.3%52.9%–55.9%
0.534
—
† Quark-G: Run on Local Ubuntu / CUDA; its step is read against a GPU re-run of Quark-H and includes the probe's input width doubling from 800 to 1600 under one fixed penalty (see the topic page).
Unless a cell is flagged, its dataset is not in that collection’s published pretraining list (the LaBraM at Scale pretraining manifest at faa387d9, checked 2026-11-05); REVE Small’s is read from the REVE corpus list (checked 2026-10-04).
Terms: LaBraM at Scale’s code and weights under MIT (Pan-Junyu/LaBraM-at-Scale @ a0b0ce27); REVE Small under the REVE Responsible Use License v1.0 (reve-small @ 3565bdcb), aggregate scientific results only. No weights are distributed.
Balanced target/nontarget sample changes the source 1:9 prevalence. One-second windows include later flashes at 0.3-second intervals; this is not an online speller estimate.
The protocol, step by step
5 participant-disjoint folds. All recordings from a person stay together. Each person contributes to the held-out predictions once.
One fixed seed (20260919); no early stopping or test-based tuning. EEGNet trains for 20 epochs per fold. Frozen encoders use training-only standardized ridge heads (alpha 100).
One fixed configuration. Foundation encoders remain frozen; small networks train from scratch. These scores do not establish optimal fine-tuned performance. No individual predictions or participant-level results are distributed.
Model terms
Spectral ridge: Trained from scratch / deterministic reference; no third-party pretrained weights. Braindecode BSD-3-Clause; MNE/scikit-learn BSD where used.
Temporal ridge: Trained from scratch / deterministic reference; no third-party pretrained weights. Braindecode BSD-3-Clause; MNE/scikit-learn BSD where used.
LaBraM: Code/repository: MIT · Checkpoint: committed in that repository; no separate weight terms
CBraMod: Code: MIT · Weights: Apache-2.0 (official model card)
EEGNet: Trained from scratch / deterministic reference; no third-party pretrained weights. Braindecode BSD-3-Clause; MNE/scikit-learn BSD where used.
XDAWNCov+TS+LR: Deterministic; no weights. pyRiemann 0.12 and scikit-learn, BSD-3-Clause.
xDAWN+LDA: Deterministic; no weights. pyRiemann 0.12 and scikit-learn, BSD-3-Clause.
ShallowFBCSPNet: Trained from scratch / deterministic reference; no third-party pretrained weights. Braindecode BSD-3-Clause; MNE/scikit-learn BSD where used.
Deep4Net: Trained from scratch / deterministic reference; no third-party pretrained weights. Braindecode BSD-3-Clause; MNE/scikit-learn BSD where used.
CSP+LDA: Trained from scratch / deterministic reference; no third-party pretrained weights. Braindecode BSD-3-Clause; MNE/scikit-learn BSD where used.
Source and licence
ds006593
CreditOpenNeuro ds006593 contributors · version 1.0.0, doi:10.18112/openneuro.ds006593.v1.0.0; original author credits retained at the linked source.
One training run per cell with the matrix seed and the matrix recipe: AdamW, 20 epochs (sleep 10), no schedule, no early stopping. This is the recipe of the EEGNet cells; it is not tuned for any network and it is not the original papers' training. Intervals describe variation across people, not across training runs.