Aggregating per-token pullback metrics via the Fréchet mean on the SPD manifold outperforms Euclidean mean pooling for sentence classification, with most of the gain attributable to geometric aggregation rather than learned encoder structure.
Barachant et al.,Pyriemann, version 0.11, 2026
2 Pith papers cite this work, alongside 8 external citations. Polarity classification is still indexing.
2
Pith papers citing it
8
external citations · OpenAlex
years
2026 2representative citing papers
ERP-XTTN uses fixed difference-wave prototypes and query-key-only cross-attention to deliver competitive leave-one-subject-out ERP classification on multiple public datasets with built-in structural interpretability.
citing papers explorer
-
Riemannian Geometry for Pre-trained Language Model Embeddings
Aggregating per-token pullback metrics via the Fréchet mean on the SPD manifold outperforms Euclidean mean pooling for sentence classification, with most of the gain attributable to geometric aggregation rather than learned encoder structure.
-
ERP-XTTN: Interpretable Prototype-Guided Cross-Attention for Cross-Subject ERP Classification
ERP-XTTN uses fixed difference-wave prototypes and query-key-only cross-attention to deliver competitive leave-one-subject-out ERP classification on multiple public datasets with built-in structural interpretability.