🙏vor visiting ⁦hessian.AI⁩ !⁦antonio vergari ⚔️ not at #ICML2025⁩ gives a fantastic talk on his recent ⁦NeurIPS Conference⁩ paper on semantic probabilistic layers in deep networks. Probabilistic circuits rock 💪arxiv.org/abs/2206.00426

Excited to present a 5-day course on "Machine Learning Beyond Static Datasets" at the European Summer School on AI in Slovenia in July! European Association for Artificial Intelligence ContinualAI Plenty of nice courses! Join to learn how to transcend static train-test all the way to lifelong learning in mine =)

🚨Are you ready!?🚨 Today we're announcing the Continual AI Unconference (CLAI Unconf) to be held in October 2023! CLAI Unconf is virtual & multi-timezone, covering diverse CL topics with pre-registered papers & contributed talks! ➡️unconf.continualai.org Please share!

great work on how #circuits and #tractable #probabilistic #models can help steer #LLMs!

🤩 “Probabilistic Circuits That Know What They Don’t Know” got accepted at #UAI2023 UAI 2025 🥳🎉 TL;DR: tractable uncertainty estimates, e.g. to detect various forms of out-of-distribution data (instead of heavy Monte-Carlo dropout) Preprint: arxiv.org/abs/2302.06544

🚨Care to distill deep classifiers' knowledge but don't have access to data & model internals? In "Deep Classifier Mimicry without Data Access" arxiv.org/abs/2306.02090 we show model-agnostic distillation without access to intermediate layers or data! 🧵👇 Steven Braun Kristian Kersting

Excited to introduce our novel dataset for visual logical learning, allowing for challenging tasks across the spectrum of AI approaches: sites.google.com/view/v-lol lukas helff Hikaru Shindo Devendra Singh Dhami | देवेंद्र सिंह धामी 🇮🇳 Kristian Kersting

🚨Announcing Self-Expanding Neural Networks (SENN): arxiv.org/abs/2307.04526 Kristian Kersting SENN can start with 1 unit in 1 layer & use natural gradients to jointly answer when/where/what to add during training. Growing on-the-fly! New cornerstone for resource-aware & adaptive ML

🙏 für das Interview! Es gäbe noch so viel mehr zu sagen. Wir brauchen den Dialog, national & international. Das Hier& Jetzt bei #KI und ihren Chancen/Risiken sind entscheidend. Wir dürfen bei der Gestaltung von KI nicht nur passiv zuschauen

Interested in how to tackle overconfidence and introduce uncertainty in #ProbabilisticCircuits? 🧠 Join our oral pres. on "Probabilistic Circuits That Know What They Don’t Know" Thu 3:15PM and chat with us at our poster (118) later at 4:30PM @ #UAI2023! 🎉

I'm featured in this great article (in 🇩🇪) on unlearning + forgetting in AI, adding perspective to the challenges & highlighting the importance of our research Computer Science, TU Darmstadt hessian.AI & ContinualAI Critical to raise the public's awareness! 🙏 Handelsblatt Larissa Holzki

Finally our novel Learning by Self-eXplaining (LSX) approach is online: arxiv.org/pdf/2309.08395… For the first time a model is trained by explaining to and receiving feedback from an internal critic, going beyond standard accuracy benefits. Check it out! :) #kerstingAIML #hkrsnd

🥳Super thrilled that my fantastic team has managed to get six #Neurips2023 papers accepted: XAI for multimodal transformer, interleaving inputs of multiple modalities & languages, semantic guidance for text2image, causality, neurosymbolic AI for RL & characteristic circuits 🙏

Wie kann ich mich schon im Studium auf KI spezialisieren? Und wie finde ich später einen Job? #TUDa Informatik-Professor Kristian Kersting Kristian Kersting Computer Science, TU Darmstadt beantwortet im Interview ZEIT Campus die wichtigsten Fragen 👇 zeit.de/campus/2023/05… via

🎉 Our work “Deep Classifier Mimicry without Data Access” w/ Steven Braun Kristian Kersting was accepted as an ✨oral✨at #AISTATS2024 arxiv.org/abs/2306.02090 🍰: data-free & model-agnostic knowledge distillation! ICLR spotlight, AIStats oral; this year is truly off to a great start

I’m sadly not at #AISTATS2024, but *please* go see Steven Braun present our work “Deep Classifier Mimicry without Data Access” on Saturday May 4th at 10:30-11:30 (oral session 8) It’s a really fun paper that shows how to distill classifiers without data access across model types 🤩

🏆🎖️Super proud supervisor moment 🥳 Steven Braun won an #AISTATS2024 Outstanding Student Paper Highlight Award🎉 His work shows how to mimic deep classifiers without access to the original data. Perfect for all the pre-trained models without data access! proceedings.mlr.press/v238/braun24b

WOW - our work "Deep Classifier Mimicry without Data Access" was awarded student paper highlight AISTATS Conference #AISTATS2024!🔥 tl;dr: we introduced model-agnostic knowledge distillation without data access! Couldn't be ANY prouder of Steven Braun & sharing the joy w/ Kristian Kersting