BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//MINERVA European Support Centre - ECPv6.17.0//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:MINERVA European Support Centre
X-ORIGINAL-URL:https://minerva4ai.eu
X-WR-CALDESC:Events for MINERVA European Support Centre
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Europe/Rome
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20250330T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20251026T010000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20260329T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20261025T010000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20270328T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20271031T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/Rome:20260421T160000
DTEND;TZID=Europe/Rome:20260421T170000
DTSTAMP:20260707T155838Z
CREATED:20260417T090849Z
LAST-MODIFIED:20260707T155838Z
UID:936-1776787200-1776790800@minerva4ai.eu
SUMMARY:MINERVA Online Journal Club #1: Large Concept Models
DESCRIPTION:Join us for our inaugural online journal club. In this session\, we will discuss together Meta AI’s work on Large Concept Models [1]. Our team from Tübingen AI Center will host and lead the discussion. The session is open to all and free of charge. We encourage you to read the paper before joining\, but it’s not a strict requirement. If you have specific questions that you want to have discussed during the session\, send them to minerva@tuebingen.ai.Large Concept Models are Meta AI’s suggestion how to improve reasoning and semantic processing of LLMs by introducing an abstraction layer in the typical LLM architecture that represents higher-level semantic units which they call “concepts”. Concepts are independent of their concrete textual representation and are thus represented using an existing sentence embedding space.  In the session\, we will discuss together what all of this means in detail.Looking forward to see many of you online and diving deep into the research trends. \n\n\n\n[1] https://ai.meta.com/research/publications/large-concept-models-language-modeling-in-a-sentence-representation-space/ \n\n\n\n\n\n\n\n\n\n\n\n\n\nRecordings
URL:https://minerva4ai.eu/event/minerva-online-journal-club-1-large-concept-models/
LOCATION:Online
ORGANIZER;CN="MINERVA":MAILTO:info@minerva4ai.eu
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Rome:20260429T143000
DTEND;TZID=Europe/Rome:20260429T153000
DTSTAMP:20260403T094453Z
CREATED:20260403T094453Z
LAST-MODIFIED:20260403T094453Z
UID:933-1777473000-1777476600@minerva4ai.eu
SUMMARY:MINERVA Webinar #4 - HPC Coding Agents
DESCRIPTION:The HPC users’ primary goal is scientific discovery. However\, many encounter continuous obstacles when running code on HPC systems. Issues such as configuring environments\, managing modules\, writing job scripts\, and defining correct paths often distract from research goals. We are working to address some of these challenges. \n\n\n\nIn this presentation\, we will introduce and demonstrate an initial coding agent developed by AMD Silo AI specifically for HPC centres. We are designing it to integrate effortlessly with popular coding assistants including Cline\, Claude Code\, and Cursor. The HPC Coding Agent is especially well-suited for scientists\, engineers\, and data teams working with AMD-based HPC systems.
URL:https://minerva4ai.eu/event/minerva-webinar-4-hpc-coding-agents/
LOCATION:Online
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Rome:20260616T100000
DTEND;TZID=Europe/Rome:20260616T110000
DTSTAMP:20260616T134629Z
CREATED:20260604T121053Z
LAST-MODIFIED:20260616T134629Z
UID:946-1781604000-1781607600@minerva4ai.eu
SUMMARY:MINERVA Webinar #5: Profiling AI workloads on GPUs – Identifying performance improvements
DESCRIPTION:The Webinar is for participants interested to learn how to identify and analyze performance improvements for AI workloads using profiling. Understand the fundamentals of profiling for AI workloads\, which can be further applied upon completion of the webinar. The webinar combines theoretical knowledge with guided coding walkthrough to help participants identify performance improvements\, understand traces\, and enhance efficiency of AI workloads on GPUs. \n\n\n\nIf you’re interested in how to use profiling to identify performance improvements for your AI workloads\, then this seminar is for you and/or if you’re working as e.g.\, a Data/AI scientist or as an engineer. \n\n\n\nAgenda \n\n\n\n\nIntroduction (5 min)\n\nWelcome & Project Introduction\n\n\n\nOverview of speaker role in AI and MINERVA\n\n\n\n\n\nIntroduction to Profiling (5 min)\n\n\n\nGetting Started with profiling (25 min)\n\n\n\nHow to profile AI workloads in practice (20min)\n\n\n\nClosing and Q&A Session (5 mins)\n\nOpen floor for questions\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nEnjoyed this event? Take our post training survey!
URL:https://minerva4ai.eu/event/minerva-webinar-5-profiling-ai-workloads-on-gpus-identifying-performance-improvements/
LOCATION:Online
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Rome:20260707T160000
DTEND;TZID=Europe/Rome:20260707T170000
DTSTAMP:20260722T100005Z
CREATED:20260702T205840Z
LAST-MODIFIED:20260722T100005Z
UID:991-1783440000-1783443600@minerva4ai.eu
SUMMARY:MINERVA Online Journal Club #2: VL-JEPA: Joint Embedding PredictiveArchitecture for Vision-language
DESCRIPTION:We’ll return with the second installment of the online journal club. In the meeting\, we will have a look together at (VL-)JEPA [1]\, a model from Yann LeCun’s research agenda around Joint Embedding Predictive Architectures. Our team from Tübingen AI Center will host and lead the discussion. The session is open to all and free of charge. We encourage you to read the paper before joining\, but it’s not a strict requirement. If you have specific questions that you want to have discussed duringthe session\, send them to minerva@tuebingen.ai. \n\n\n\n(VL-)JEPA [1] is an approach to enhance vision-language models by introducing a predictive architecture. Instead of generating text tokens directly\, it predicts continuous embeddings representing abstractsemantic concepts. The intuition behind this is to have the model to focus on task-relevant meaning\, abstracting from surface-level linguistic variations\, leading to more efficient and semantically rich representations. \n\n\n\nLooking forward to see many of you online and diving deep into the research trends. \n\n\n\n[1] https://openreview.net/forum?id=tjimrqc2BU \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\nRecordings
URL:https://minerva4ai.eu/event/minerva-online-journal-club-2-vl-jepa-joint-embedding-predictivearchitecture-for-vision-language/
LOCATION:Online
ORGANIZER;CN="MINERVA":MAILTO:info@minerva4ai.eu
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=Europe/Rome:20260804T160000
DTEND;TZID=Europe/Rome:20260804T170000
DTSTAMP:20260722T100257Z
CREATED:20260722T095659Z
LAST-MODIFIED:20260722T100257Z
UID:1034-1785859200-1785862800@minerva4ai.eu
SUMMARY:MINERVA Online Journal Club #3: Exploring Mamba-3
DESCRIPTION:The third instalment of the online journal club will feature the third instalment of the Mamba model series. We will discuss and dissect the ICLR 2026 paper “Mamba-3: Improved Sequence Modeling using State Space Principles” by Lahoti et al. [1]. Again\, the Tübingen AI Center team will host and lead the discussion. The session is open to all and free of charge. We encourage you to read the paper before joining\, but it’s not a strict requirement. If you have specific questions that you want to have discussed during the session\, send them to minerva@tuebingen.ai. \n\n\n\nMamba-3 [1] is the third instalment of an alternate LLM model architecture [2\, 3]. While the majority of transformer-based LLM architectures are structured as auto-regressive models\, mapping sequences of length n to sequences of length n+1\, models in the Mamba family capture the state of the model in a latent state\, thus circumventing the linear memory and quadratic compute cost at inference time. We will have a close look what implications this has both in terms of computational requirements\, but also in model performance and what future improvements might be expected from models in this family. \n\n\n\nLooking forward to see many of you online and diving deep into the research trends.  \n\n\n\n[1] https://openreview.net/forum?id=HwCvaJOiCj[2] https://openreview.net/forum?id=ztn8FCR1td[3] https://openreview.net/forum?id=tEYskw1VY2
URL:https://minerva4ai.eu/event/minerva-online-journal-club-3-mamba-3/
ORGANIZER;CN="T%C3%BCbingen AI Center":MAILTO:minerva@tuebingen.ai
END:VEVENT
END:VCALENDAR