The paper reports the winning solution to the MoCha 2026 Parkinsonian Gait Benchmark, achieving a macro‑F1 score of 0.6945 on unseen clinical sites. The approach relies on a frozen public motion encoder followed by a single 4×512 linear layer, and gains are largely attributed to three key steps: exact replication of the benchmark’s head recipe, averaging per‑walk posteriors at the subject level, and a label‑free transductive calibration of feature means and decision thresholds. Extensive ablation studies show that fine‑tuning the encoder or using alternative encoders does not improve performance, and the subject‑level aggregation is identified as the primary contributor to the top score.
arXiv:2608.23730v1 Announce Type: new Abstract: We grade MDS-UPDRS gait severity from SMPL motion using three frozen MotionAGFormer encoders as featurizers, reaching macro-F1 0.58 on a hidden, multi-...
arXiv:2605. 15862v2 Announce Type: replace Abstract: Understanding adaptive biomechanical systems requires distinguishing observable performance, static multivariate representation, longitudinal displacement, and internal approximation of observed change.
arXiv:2606. 25270v1 Announce Type: new Abstract: Keystroke dynamics have been explored extensively as a passive digital biomarker for Parkinson's disease (PD), typically by extracting summary statistics from typing timing and training a classifier to discriminate PD from healthy controls.
arXiv:2605. 00778v2 Announce Type: replace Abstract: In biomechanical systems, observable performance is often used as a proxy for underlying organization, although similar outputs may arise from different adaptive configurations.
arXiv:2607. 27565v1 Announce Type: new Abstract: Pattern-recognition control promises a myoelectric prosthesis that responds to many intended gestures rather than one or two, but the promise has stayed in the laboratory.
Pattern-recognition control promises a myoelectric prosthesis that responds to many intended gestures rather than one or two, but the promise has stayed in the laboratory. A recogniser trained on one person rarely transfers to the next, and useful performance usually demands a fresh round of labelled calibration from the end user.
arXiv:2607. 24519v2 Announce Type: replace Abstract: Pretrained EEG foundation models are proposed for clinical decoding, but whether reported gains transfer across populations or survive negative controls is unclear.
arXiv:2606. 24960v1 Announce Type: new Abstract: Tailoring stroke rehabilitation requires assessing how movements are organized, not merely if they succeed.
arXiv:2608. 06122v1 Announce Type: cross Abstract: Inspired by recent evidence that transformer architectures benefit from Self-PreTraining (SPT) on long-context benchmarks, we investigate whether similar gains extend to multimodal, multivariate, and even simple univariate medical time series.
The paper critiques the common practice of evaluating specialist language models solely by endpoint gains, arguing that the underlying model updates are largely unexamined. It introduces a paired weight‑delta path audit and applies it to two public generalist‑to‑medical‑specialist checkpoint pairs, showing that the decoder‑side updates largely reconstruct the observed benchmark improvements. However, the audit finds that the improvements are not neatly localized to a single component family, indicating that component‑level explanations are more complex than previously assumed.
arXiv:2605. 10840v3 Announce Type: replace-cross Abstract: We present Clin-JEPA, a multi-phase co-training framework for joint-embedding predictive (JEPA) pretraining on EHR patient trajectories.
arXiv:2606. 24394v1 Announce Type: cross Abstract: Electroencephalography (EEG) is the dominant non-invasive modality for brain-computer interfaces (BCIs), yet reliable decoding of motor imagery is hampered by inter- and intra-individual variability.