{"ID":2874859,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2509.02982","arxiv_id":"2509.02982","title":"StableSleep: Source-Free Test-Time Adaptation for Sleep Staging with Lightweight Safety Rails","abstract":"Sleep staging models often degrade when deployed on patients with unseen physiology or recording conditions. We propose a streaming, source-free test-time adaptation (TTA) recipe that combines entropy minimization (Tent) with Batch-Norm statistic refresh and two safety rails: an entropy gate to pause adaptation on uncertain windows and an EMA-based reset to reel back drift. On Sleep-EDF Expanded, using single-lead EEG (Fpz-Cz, 100 Hz, 30s epochs; R\u0026K to AASM mapping), we show consistent gains over a frozen baseline at seconds-level latency and minimal memory, reporting per-stage metrics and Cohen's k. The method is model-agnostic, requires no source data or patient calibration, and is practical for on-device or bedside use.","short_abstract":"Sleep staging models often degrade when deployed on patients with unseen physiology or recording conditions. We propose a streaming, source-free test-time adaptation (TTA) recipe that combines entropy minimization (Tent) with Batch-Norm statistic refresh and two safety rails: an entropy gate to pause adaptation on unce...","url_abs":"https://arxiv.org/abs/2509.02982","url_pdf":"https://arxiv.org/pdf/2509.02982v1","authors":"[\"Hritik Arasu\",\"Faisal R Jahangiri\"]","published":"2025-09-03T03:42:31Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\",\"cs.NE\",\"q-bio.NC\"]","methods":"[]","has_code":false}