BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Embed2Scale - ECPv6.17.0//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-ORIGINAL-URL:https://embed2scale.eu
X-WR-CALDESC:Events for Embed2Scale
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:UTC
BEGIN:STANDARD
TZOFFSETFROM:+0000
TZOFFSETTO:+0000
TZNAME:UTC
DTSTART:20250101T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=UTC:20260424T150000
DTEND;TZID=UTC:20260424T160000
DTSTAMP:20260429T090814Z
CREATED:20260212T160142Z
LAST-MODIFIED:20260429T090814Z
UID:10000022-1777042800-1777046400@embed2scale.eu
SUMMARY:Webinar: "Embedding Space Metrics Constrained Fine-Tuning"
DESCRIPTION:Home » Event\n\n\n\n\n\n\n\n\n\n\n\n\nEnhancing maritime domain awareness using AI and embeddings\n\n\n\n\n\n\n\nCEST \n\n\n\n\n\n\n\n\nWhile fine-tuning Geospatial Foundation Models (GFMs) has become the standard for domain adaptation\, it often leads to catastrophic forgetting or ‘representation drift\,’ where the model’s fundamental understanding of semantic relationships is warped. This webinar introduces Embedding Space Metrics Constrained Fine-Tuning (ESM-CFT)\, an optimisation method in the TerraTorch framework designed to preserve the structural integrity of the latent space during supervised fine-tuning. \n\n\n\nUnlike traditional methods that rely solely on cross-entropy loss\, ESM-CFT incorporates a geometric constraint based on distance metrics to ensure that the relative positioning of representations in the embedding space remains stable. \n\n\n\nThe webinar presents promising preliminary experimental results and opens the stage for discussion of future work. \n\n\n\nWatch the webinar
URL:https://embed2scale.eu/event/webinar-embedding-space-metrics-constrained-fine-tuning/
CATEGORIES:Event
ATTACH;FMTTYPE=image/jpeg:https://embed2scale.eu/wp-content/uploads/2026/02/Embedding-Space-Metrics-Constrained-Fine-Tuning_Web-1-scaled.jpg
END:VEVENT
END:VCALENDAR