BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Huntsville AI - ECPv6.8.3//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-ORIGINAL-URL:https://www.hsv.ai
X-WR-CALDESC:Events for Huntsville AI
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/Chicago
BEGIN:DAYLIGHT
TZOFFSETFROM:-0600
TZOFFSETTO:-0500
TZNAME:CDT
DTSTART:20250309T080000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0500
TZOFFSETTO:-0600
TZNAME:CST
DTSTART:20251102T070000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/Chicago:20251210T180000
DTEND;TZID=America/Chicago:20251210T193000
DTSTAMP:20260422T225008
CREATED:20251204T140904Z
LAST-MODIFIED:20251208T033805Z
UID:2066-1765389600-1765395000@www.hsv.ai
SUMMARY:Virtual Paper Review - AI World Models
DESCRIPTION:Join us for our monthly virtual Paper Review as we explore the multifaceted world of World Models (WMs)! This session will utilize a recent survey paper to help explore a spectrum of uses for these models and connect them to some of the other topics we have covered in this series such as reasoning\, multimodal embeddings\, and diffusion transformers. \nWe will analyze the current state of world models\, who are the big players\, what are they doing? And where might they be taking us in 2026? \nHigh Level Topics : \n\nConceptual Links: Connecting modern WMs to the psychological concept of Mental Models and engineering concepts like Digital Twins.\nRepresentation & Recovery: Discussing Implicit Representation and World Model Recovery.\nFuture Predictions: An overview of prediction methodologies.\nDecision-Making: Advances in Model-based Reinforcement Learning (state transition dynamics) and the integration of LLMs/MLLMs for planning and action generation.\nWorld Knowledge: Categorizing the knowledge implicitly held within WMs: Global Physical World\, Local Physical World (cognitive maps)\, and Human Society (Theory of Mind).\nGenerative Progress: The evolution from video generation (Sora) to fully Embodied Environments (Genie3 / Marble).\nCritical Applications: Reviewing the requirements for WMs in areas such as Game\, Embodied\, Urban\, and Societal Intelligence.\nOpen Research Lanes: Identifying key unsolved research challenges.\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nLinks: \n\nV-JEPA 2: Self-Supervised Video Models Enable Understanding\, Prediction and Planning – https://arxiv.org/pdf/2506.09985\nLeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristics – https://arxiv.org/pdf/2511.08544v3\n\nDetails: \n\nDate – 12/10/2025\nTime – 6-7:30pm\nLocation – VIRTUAL\nGoogle Meet – https://meet.google.com/bry-ruop-vpq\nDial in:‪ (US) +1 650-667-3474‬ PIN: ‪989 475 586‬#
URL:https://www.hsv.ai/event/virtual-lepaper-review/
ATTACH;FMTTYPE=image/png:https://www.hsv.ai/wp-content/uploads/2025/12/jp-ai-world-models.png
END:VEVENT
END:VCALENDAR