Paul-Vincent Roll (he/him)<p>Mhhhhhhhhh 🙈😄 Do I continue to investigate this? 🧐</p><p>This is using the on device foundation models to query information from a relationship menu and answer questions about it. I don't know yet how helpful this is haha probably not much. <a href="https://whisper.tf/tags/foundationmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>foundationmodels</span></a> <a href="https://whisper.tf/tags/swift" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>swift</span></a></p>
eupolicy.social is one of the many independent Mastodon servers you can use to participate in the fediverse.
This Mastodon server is a friendly and respectful discussion space for people working in areas related to EU policy. When you request to create an account, please tell us something about you.
Administered by:
Server stats:
201active users
eupolicy.social: About · Profiles directory · Privacy policy
Mastodon: About · Get the app · Keyboard shortcuts · View source code · v4.4.3
#foundationmodels
2 posts · 2 participants · 1 post today
Taylor Arndt<p>Working on something exciting: Perspective Chat.<br>It is similar to ChatGPT but built on Apple’s Foundation Models. It remembers tasks, applies priorities, is designed for accessibility, and will support MCP integration.<br>Perspective AT is also close to release.<br><a href="https://techopolis.social/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://techopolis.social/tags/Accessibility" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Accessibility</span></a> <a href="https://techopolis.social/tags/TechForGood" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TechForGood</span></a> <a href="https://techopolis.social/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> <a href="https://techopolis.social/tags/OpenSource" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>OpenSource</span></a></p>
John Ellenich<p>Kinda fun feature I added to <span class="h-card" translate="no"><a href="https://indieapps.space/@remainders" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>remainders</span></a></span> using <a href="https://techhub.social/tags/SwiftUI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SwiftUI</span></a> and Apple’s on device <a href="https://techhub.social/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> in <a href="https://techhub.social/tags/iOS26" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>iOS26</span></a>…</p><p>I take whatever the user titled the event and pass it into the on device LLM (fast, free, no Internet connection required 🎉), have it generate 3-4 “concepts” based on the title, pass the concepts as an array into the in-app, Image Playground sheet, and generate a cover image for their event.</p><p>It’ll even do proper nouns and pull those people into the image (if you’ve set that up).</p>
TelemetryDeck<p>Are you planning to ship a new app or update next week for <a href="https://social.telemetrydeck.com/tags/iOS26" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>iOS26</span></a> with <a href="https://social.telemetrydeck.com/tags/LiquidGlass" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LiquidGlass</span></a> support or using new frameworks like <a href="https://social.telemetrydeck.com/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a>? ✨</p><p>Then let us know! We're here to support <a href="https://social.telemetrydeck.com/tags/IndieDevs" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>IndieDevs</span></a> and will like + share your post. 🙌</p><p>Just comment what's new below & follow us: 👇 <br><a href="https://social.telemetrydeck.com/tags/iOSDev" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>iOSDev</span></a> <a href="https://social.telemetrydeck.com/tags/AppleEvent" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AppleEvent</span></a> <a href="https://social.telemetrydeck.com/tags/macOS26" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>macOS26</span></a> <a href="https://social.telemetrydeck.com/tags/visionOS" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>visionOS</span></a></p>
eicker.news ᳇ tech news<p><a href="https://eicker.news/tags/Meta" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Meta</span></a> is reorganising its <a href="https://eicker.news/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://eicker.news/tags/organisation" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>organisation</span></a>, again: this time into four new groups under the <a href="https://eicker.news/tags/MetaSuperintelligenceLabs" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MetaSuperintelligenceLabs</span></a> (<a href="https://eicker.news/tags/MSL" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MSL</span></a>). The new groups will focus on <a href="https://eicker.news/tags/foundationmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>foundationmodels</span></a>, <a href="https://eicker.news/tags/research" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>research</span></a>, <a href="https://eicker.news/tags/productintegration" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>productintegration</span></a>, and <a href="https://eicker.news/tags/infrastructure" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>infrastructure</span></a>. <a href="https://techcrunch.com/2025/08/19/meta-is-shaking-up-its-ai-org-again/?eicker.news" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">techcrunch.com/2025/08/19/meta</span><span class="invisible">-is-shaking-up-its-ai-org-again/?eicker.news</span></a> <a href="https://eicker.news/tags/tech" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>tech</span></a> <a href="https://eicker.news/tags/media" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>media</span></a> <a href="https://eicker.news/tags/news" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>news</span></a></p>
Donovan H<p>A glimpse at a little work in progress - trying to set up a nice set of demo apps showing how the new <a href="https://mastodon.ie/tags/swift" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>swift</span></a> <a href="https://mastodon.ie/tags/foundationmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>foundationmodels</span></a> work. 20 demos so far - lots of small UI issues but they all work to some extent. If all goes well I’ll create some blogs to explain how to build them 🎉</p>
apfeltalk :verified:<p>Apple verliert weiter KI-Expert:innen an Meta: Ein analytischer Überblick<br>Apple erlebt einen kontinuierlichen Abgang von KI-Expert:innen, die verstärkt zum Konkurrenten Met<br><a href="https://www.apfeltalk.de/magazin/news/apple-verliert-weiter-ki-expertinnen-an-meta-ein-analytischer-ueberblick/" rel="nofollow noopener" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">apfeltalk.de/magazin/news/appl</span><span class="invisible">e-verliert-weiter-ki-expertinnen-an-meta-ein-analytischer-ueberblick/</span></a><br><a href="https://creators.social/tags/KI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>KI</span></a> <a href="https://creators.social/tags/News" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>News</span></a> <a href="https://creators.social/tags/Abwerbung" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Abwerbung</span></a> <a href="https://creators.social/tags/AIEntwicklung" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AIEntwicklung</span></a> <a href="https://creators.social/tags/Apple" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Apple</span></a> <a href="https://creators.social/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> <a href="https://creators.social/tags/Innovation" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Innovation</span></a> <a href="https://creators.social/tags/KITeams" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>KITeams</span></a> <a href="https://creators.social/tags/KnstlicheIntelligenz" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>KnstlicheIntelligenz</span></a> <a href="https://creators.social/tags/LargeLanguageModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LargeLanguageModels</span></a> <a href="https://creators.social/tags/Meta" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Meta</span></a> <a href="https://creators.social/tags/Mitarbeiterbindung" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Mitarbeiterbindung</span></a> <a href="https://creators.social/tags/Personalwechsel" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Personalwechsel</span></a> <a href="https://creators.social/tags/Siri" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Siri</span></a> <a href="https://creators.social/tags/Technologiewettbewerb" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Technologiewettbewerb</span></a> <a href="https://creators.social/tags/Vergtung" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Vergtung</span></a></p>
Europe Says<p><a href="https://www.europesays.com/2255431/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="">europesays.com/2255431/</span><span class="invisible"></span></a> AI models with systemic risks given pointers on how to comply with EU AI rules <a href="https://pubeurope.com/tags/Europe" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Europe</span></a> <a href="https://pubeurope.com/tags/EuropeanUnion" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>EuropeanUnion</span></a> <a href="https://pubeurope.com/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> <a href="https://pubeurope.com/tags/models" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>models</span></a> <a href="https://pubeurope.com/tags/SystemicRisks" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SystemicRisks</span></a> <a href="https://pubeurope.com/tags/TheEuropeanCommission" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TheEuropeanCommission</span></a></p>
Victoria Stuart 🇨🇦 🏳️⚧️<p>[Science] Researchers claim their AI model simulates the human mind. Others are skeptical<br><a href="https://www.science.org/content/article/researchers-claim-their-ai-model-simulates-human-mind-others-are-skeptical" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">science.org/content/article/re</span><span class="invisible">searchers-claim-their-ai-model-simulates-human-mind-others-are-skeptical</span></a></p><p>[Nature] A foundation model to predict and capture human cognition<br><a href="https://www.nature.com/articles/s41586-025-09215-4" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="ellipsis">nature.com/articles/s41586-025</span><span class="invisible">-09215-4</span></a></p><p><a href="https://mastodon.social/tags/ML" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ML</span></a> <a href="https://mastodon.social/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://mastodon.social/tags/cognition" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>cognition</span></a> <a href="https://mastodon.social/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> <a href="https://mastodon.social/tags/CognitiveModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>CognitiveModels</span></a></p>
:rss: Qiita - 人気の記事<p>【WWDC25】iPhoneだけでLLM⁉︎ Foundation Modelsを“3行”で動かしてみた<br><a href="https://qiita.com/takatein/items/fa28aa4ace902fd7a7cc?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">qiita.com/takatein/items/fa28a</span><span class="invisible">a4ace902fd7a7cc?utm_campaign=popular_items&utm_medium=feed&utm_source=popular_items</span></a></p><p><a href="https://rss-mstdn.studiofreesia.com/tags/qiita" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>qiita</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/AppleIntelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AppleIntelligence</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/iOS26" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>iOS26</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/WWDC2025" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>WWDC2025</span></a> <a href="https://rss-mstdn.studiofreesia.com/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a></p>
Jaanus Kase<p><a href="https://iosdev.space/tags/wwdc25" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>wwdc25</span></a> <a href="https://iosdev.space/tags/Xcode" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Xcode</span></a> <a href="https://iosdev.space/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://iosdev.space/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> <a href="https://iosdev.space/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a></p><p>If you want to use Xcode AI features, OR if you want to use Foundation Models AI API on iOS Simulator or macOS, looks like the only supported configuration is having Tahoe beta on the internal SSD of your Apple Silicon Mac</p><p>I sort of get why, see the thread (tl;dr: probably performance), but is there any official info from Apple about this requirement? Nothing in the few WWDC videos I saw so far<br><a href="https://mastodon.social/@felix_schwarz/114670973442659374" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">mastodon.social/@felix_schwarz</span><span class="invisible">/114670973442659374</span></a></p>
Jaanus Kase<p><a href="https://iosdev.space/tags/wwdc25" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>wwdc25</span></a> <a href="https://iosdev.space/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> <a href="https://iosdev.space/tags/Xcode" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Xcode</span></a> <a href="https://iosdev.space/tags/LLM" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>LLM</span></a> </p><p>Has anyone seen the Foundation Models actually working? Am I holding it wrong? I am not getting any responses either in the playground or app, and there is a bunch of this crap in the console when running the app.</p><p>Running Xcode 26 beta on macOS 26 beta.</p><p>Are there regional restrictions? Xcode AI chat also is “unavailable in my region”.</p>
KINEWS24<p>🚀 Apple 2025: KI direkt aufs Gerät! Was steckt dahinter?</p><p>• Neue Foundation Models enthüllt<br>• Datenschutz + Power auf iPhones<br>• Entwickler-Tools für smarte Apps</p><p><a href="https://mastodon.social/tags/ai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ai</span></a> <a href="https://mastodon.social/tags/ki" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ki</span></a> <a href="https://mastodon.social/tags/artificialintelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>artificialintelligence</span></a> <a href="https://mastodon.social/tags/Apple" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Apple</span></a> <a href="https://mastodon.social/tags/WWDC25" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>WWDC25</span></a> <a href="https://mastodon.social/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a></p><p>Jetzt LIKEN, teilen, LESEN und FOLGEN! Schreib uns in den Kommentaren!</p><p><a href="https://kinews24.de/apples-wwdc-2025-foundation-models-on-device-ki/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">kinews24.de/apples-wwdc-2025-f</span><span class="invisible">oundation-models-on-device-ki/</span></a></p>
Cyberagentur<p>ChatGPT & Co für sicherheitsrelevante Einsatzfelder nutzbar machen <br>Die <span class="h-card" translate="no"><a href="https://social.bund.de/@Cyberagentur" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>Cyberagentur</span></a></span> hat <a href="https://social.bund.de/tags/Forschungsprogramm" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Forschungsprogramm</span></a> <a href="https://social.bund.de/tags/HEGEMON" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HEGEMON</span></a> zur Bewertung und Anpassung generativer <a href="https://social.bund.de/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> für sicherheitskritische Anwendungen gestartet.<br>Entwicklung neuer Benchmarks für KI-Modelle und Anwendung auf komplexe Aufgaben aus dem Geoinformationswesen. <br>Mehr Infos zur Vergabe: <a href="https://t1p.de/qcuq4" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">t1p.de/qcuq4</span><span class="invisible"></span></a><br><a href="https://social.bund.de/tags/Cyberagentur" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Cyberagentur</span></a> <a href="https://social.bund.de/tags/HEGEMON" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HEGEMON</span></a> <a href="https://social.bund.de/tags/KI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>KI</span></a> <a href="https://social.bund.de/tags/Sicherheit" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Sicherheit</span></a> <a href="https://social.bund.de/tags/Benchmarking" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Benchmarking</span></a> <a href="https://social.bund.de/tags/PCP" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>PCP</span></a> <a href="https://social.bund.de/tags/Geoinformation" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Geoinformation</span></a> <a href="https://social.bund.de/tags/Forschung" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Forschung</span></a> <a href="https://social.bund.de/tags/KITransparenz" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>KITransparenz</span></a></p>
UK<p><a href="https://www.europesays.com/uk/135577/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://www.</span><span class="">europesays.com/uk/135577/</span><span class="invisible"></span></a> One of Europe’s top AI researchers raised a $13M seed to crack the ‘holy grail’ of models <a href="https://pubeurope.com/tags/3D" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>3D</span></a> <a href="https://pubeurope.com/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> <a href="https://pubeurope.com/tags/ArtificialIntelligence" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ArtificialIntelligence</span></a> <a href="https://pubeurope.com/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> <a href="https://pubeurope.com/tags/munich" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>munich</span></a> <a href="https://pubeurope.com/tags/SpAItial" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>SpAItial</span></a> <a href="https://pubeurope.com/tags/Technology" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Technology</span></a> <a href="https://pubeurope.com/tags/UK" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>UK</span></a> <a href="https://pubeurope.com/tags/UnitedKingdom" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>UnitedKingdom</span></a></p>
DurstewitzLab<p>Can time series (TS) <a href="https://mathstodon.xyz/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> (FM) like Chronos zero-shot generalize to unseen <a href="https://mathstodon.xyz/tags/DynamicalSystems" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>DynamicalSystems</span></a> (DS)? <a href="https://mathstodon.xyz/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a></p><p>No, they cannot!</p><p>But *DynaMix* can, the first TS/DS foundation model based on principles of DS reconstruction, capturing the long-term evolution of out-of-domain DS: <a href="https://arxiv.org/pdf/2505.13192v1" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">arxiv.org/pdf/2505.13192v1</span><span class="invisible"></span></a></p><p>Unlike TS foundation models, DynaMix exhibits <a href="https://mathstodon.xyz/tags/ZeroShotLearning" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>ZeroShotLearning</span></a> of long-term stats of unseen DS, incl. attractor geometry & power spectrum, w/o *any* re-training, just from a context signal. <br>It does so with only 0.1% of the parameters of Chronos & 10x faster inference times than the closest competitor.</p><p>It often even outperforms TS FMs on forecasting diverse empirical time series, like weather, traffic, or medical data, typically used to train TS FMs. <br>This is surprising, cos DynaMix’ training corpus consists *solely* of simulated limit cycles & chaotic systems, no empirical data at all!</p><p>And no, it’s neither based on Transformers nor Mamba – it’s a new type of mixture-of-experts architecture based on the recently introduced AL-RNN (<a href="https://proceedings.neurips.cc/paper_files/paper/2024/file/40cf27290cc2bd98a428b567ba25075c-Paper-Conference.pdf" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">proceedings.neurips.cc/paper_f</span><span class="invisible">iles/paper/2024/file/40cf27290cc2bd98a428b567ba25075c-Paper-Conference.pdf</span></a>), specifically trained for DS reconstruction.</p><p>Remarkably, DynaMix not only generalizes zero-shot to novel DS, but it can even generalize to new initial conditions and regions of state space not covered by the in-context information.</p><p>We dive a bit into the reasons why current time series FMs not trained for DS reconstruction fail, and conclude that a DS perspective on time series forecasting & models may help to advance the <a href="https://mathstodon.xyz/tags/TimeSeriesAnalysis" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TimeSeriesAnalysis</span></a> field.</p>
Helmholtz Imaging<p>The <a href="https://helmholtz.social/tags/HFMI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HFMI</span></a> ELLIS Workshop on <a href="https://helmholtz.social/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> in Science brought together top <a href="https://helmholtz.social/tags/AI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>AI</span></a> & research minds in Berlin to explore the power of foundation models across scientific domains.</p><p>We're proud to be part of the Helmholtz Foundation Model Initiative driving this forward.</p><p>More 👉 <a href="https://bit.ly/AI-4-Science" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">bit.ly/AI-4-Science</span><span class="invisible"></span></a></p><p><span class="h-card" translate="no"><a href="https://helmholtz.social/@association" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>association</span></a></span> <span class="h-card" translate="no"><a href="https://social.bund.de/@MDC_Berlin" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>MDC_Berlin</span></a></span> </p><p><a href="https://helmholtz.social/tags/HFMI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HFMI</span></a> </p><p>Image: Svea Pietschmann, Max Delbrück Center</p>
Anita Graser 🇪🇺🇺🇦🇬🇪<p>The quest for a fair TimeGPT benchmark</p><p>At the end of yesterday's <a href="https://fosstodon.org/tags/TimeGPT" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TimeGPT</span></a> for mobility post, we concluded that TimeGPT's trainingset probably included a copy of the popular BikeNYC timeseries dataset and that, therefore, we were not looking at a fair comparison ...</p><p><a href="http://anitagraser.com/2025/03/29/the-quest-for-a-fair-timegpt-benchmark/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">http://</span><span class="ellipsis">anitagraser.com/2025/03/29/the</span><span class="invisible">-quest-for-a-fair-timegpt-benchmark/</span></a></p><p><a href="https://fosstodon.org/tags/GeoAI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GeoAI</span></a> <a href="https://fosstodon.org/tags/MobilityDataScience" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MobilityDataScience</span></a> <a href="https://fosstodon.org/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a> <a href="https://fosstodon.org/tags/GISChat" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>GISChat</span></a></p>
Anita Graser 🇪🇺🇺🇦🇬🇪<p><a href="https://fosstodon.org/tags/TimeGPT" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>TimeGPT</span></a> for <a href="https://fosstodon.org/tags/mobility" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>mobility</span></a>: Can foundation models outperform classic machine learning models for flow predictions?</p><p>tldr; Maybe. Preliminary results certainly are impressive. </p><p><a href="http://anitagraser.com/2025/03/28/can-foundation-models-outperform-classic-machine-learning-models-for-mobility-predictions/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">http://</span><span class="ellipsis">anitagraser.com/2025/03/28/can</span><span class="invisible">-foundation-models-outperform-classic-machine-learning-models-for-mobility-predictions/</span></a></p><p><a href="https://fosstodon.org/tags/MobilityDataScience" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>MobilityDataScience</span></a> <a href="https://fosstodon.org/tags/gischat" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>gischat</span></a> <a href="https://fosstodon.org/tags/urbanmobility" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>urbanmobility</span></a> <a href="https://fosstodon.org/tags/geoai" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>geoai</span></a> <a href="https://fosstodon.org/tags/foundationmodels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>foundationmodels</span></a></p>
Helmholtz Imaging<p>We're excited to be part of this lighthouse program of <span class="h-card" translate="no"><a href="https://helmholtz.social/@association" class="u-url mention" rel="nofollow noopener" target="_blank">@<span>association</span></a></span>, (co-)leading three <a href="https://helmholtz.social/tags/HFMI" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HFMI</span></a> projects: AqQua, The Human Radiome Project and the Synergy Unit.</p><p>𝗠𝗼𝗿𝗲<br>🔗 HFMI: <a href="https://hfmi.helmholtz.de/" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">hfmi.helmholtz.de/</span><span class="invisible"></span></a> <br>🔗 AqQua: <a href="https://bit.ly/AqQua" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">bit.ly/AqQua</span><span class="invisible"></span></a> <br>🔗 The Human Radiome Project & Synergy Unit: <a href="https://bit.ly/THRP-SynergyUnit" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="">bit.ly/THRP-SynergyUnit</span><span class="invisible"></span></a> </p><p><a href="https://helmholtz.social/tags/Helmholtz" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>Helmholtz</span></a> <a href="https://helmholtz.social/tags/HelmholtzImaging" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>HelmholtzImaging</span></a> <a href="https://helmholtz.social/tags/FoundationModels" class="mention hashtag" rel="nofollow noopener" target="_blank">#<span>FoundationModels</span></a></p><p><a href="https://helmholtz.social/@association/113962295209701941" rel="nofollow noopener" translate="no" target="_blank"><span class="invisible">https://</span><span class="ellipsis">helmholtz.social/@association/</span><span class="invisible">113962295209701941</span></a></p>
TrendingLive feeds
Mastodon is the best way to keep up with what's happening.
Follow anyone across the fediverse and see it all in chronological order. No algorithms, ads, or clickbait in sight.
Create accountLoginDrag & drop to upload