AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

PATCHCODE: Discrete Latent Predictive Learning for EEG Foundation Model

KIEREN YU, Ziyang LIU, Chang Huang, Kaishun Wu

ICML 2026 · Label-efficient and robust learning › Recognition and long-tail · approach: Dense features and foundation priors

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add C2P pathway (ablation)PatchCode · SEED-Vaccuracy: .407 -> .414Improved
add P2C pathway (ablation)PatchCode · SEED-Vaccuracy: .410 -> .414Improved
add P2P pathway (ablation)PatchCode · SEED-Vaccuracy: .412 -> .414Improved
add C2P pathway (ablation)PatchCode · HMCaccuracy: .742 -> .756Improved
add P2C pathway (ablation)PatchCode · HMCaccuracy: .745 -> .756Improved
add P2P pathway (ablation)PatchCode · HMCaccuracy: .748 -> .756Improved
add C2P pathway (ablation)PatchCode · BCI-IV-2aaccuracy: .572 -> .584Improved
add P2C pathway (ablation)PatchCode · BCI-IV-2aaccuracy: .574 -> .584Improved
add P2P pathway (ablation)PatchCode · BCI-IV-2aaccuracy: .578 -> .584Improved
scale model from Tiny to LargePatchCode · SEED-Vbalanced accuracy: 0.4251 -> 0.4346Improved
add region bias attention (ablation)PatchCode · avg of ablation datasets (Appendix)average score: 0.7362 -> 0.7473Improved
add region regularizer lambda_reg=0.1 (ablation)PatchCode tokenizer · avg of ablation datasets (Appendix)average score: 0.7340 -> 0.7473Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy