ETH AI Center

We’re now on Bluesky!🦋 Follow us at @ eth-ai-center.bsky.social (bsky.app/profile/eth-ai…) for updates on AI research, events, and more. See you there! #AI #ETHZurich #BLUESKY

Generative models are great at mimicking data — but real (scientific) discovery requires going beyond it. Excited to present our paper “Provable Maximum Entropy Manifold Exploration via Diffusion Models” this Wednesday at ICML 2025! We propose a scalable, theoretically grounded…

Great energy at RobotX Innovation Day @ETH_en — exciting demos and great discussions with the robotics community! 🔥 Our faculty and fellows are sharing some of their best work today! 🤖

🔥 Happening now @ETH_en, RobotX Innovation Day is attracting huge attention from the robotics community! 🎉 🤖 Here is a quick walk-around: @leggedrobotics @crl_ethz @cvg_ethz @ASL_ETHZ @srl_ethz @eth_dmavt @CSatETH @ETH_AI_Center @duatic_ag @FlexionRobotics @mimicrobotics …

The paper submission deadline is within 1 week: June 21st, 2025, AoE The challenge deadline is in 2 weeks: July 1st, 2025, AoE 📢Call for Papers: sites.google.com/view/ai4mathwo… 🎯Challenge 1, File-level Automated Proof Engineering (APE) of Formal Math Libraries (APE-Bench I):…

Register now (first-come first-served) for the "Math of Trustworthy ML workshop" at #LagoMaggiore, Switzerland, Oct 12-16 this year, with a great speaker lineup and the opportunity to present your work as a poster session or contributed talk. Details @ mmlworkshop.ethz.ch

🚨Time to push LLMs further! 📚LEXam: The legal reasoning benchmark you’ve been waiting for: • 340 exams, 116 courses (EN/DE) • Long-form, process-focused questions • Expert-level LLM Judges • Rich meta for targeted diagnostics • Contamination-proof, extendable MCQs [1/6]🧵

🎉 Today, @dmitrievdaniil7, our second PhD Fellow of the @ETH_AI_Center, graduated! He was supervised by ETH AI Center Faculty members Prof. @FannyYangETH and Prof. @AfonsoSBandeira. Congratulations, Dr. Daniil Dmitriev! Next, he will join @Penn as a Postdoctoral Researcher.

Excited to present the 𝗳𝗶𝗿𝘀𝘁 𝗽𝗮𝗽𝗲𝗿 𝗼𝗳 𝗺𝘆 𝗣𝗵𝗗 ! 🎉 Let's chat if you're attending #ICRA2025, I'll be presenting: 📍Mon, May 19 | 9:40-10:40 — Workshop on Field Robotics 📍Thu, May 22 | 08:30–09:00 — Session ThAT5 @ETH @ETH_AI_Center

Check out our #ICRA2025 paper ! Robust, precise, and terrain-aware—our RL controller significantly improves over baselines for whole-body 6-DoF tracking.💥 Project website: leggedrobotics.github.io/wholebody-pose… Work led by: @TifannyPortela, Andrei Cramariuc, Mayank Mittal and Marco Hutter

Presenting 2 posters today at ICLR. Come check them out! 10am ➡️ #502: Scalable Extraction of Training Data from Aligned, Production Language Models 3pm ➡️ #324: Adversarial Perturbations Cannot Reliably Protect Artists From Generative AI

ZurichCV#9 will be on Tuesday, April 29th at the @ETH_AI_Center! Linus Scheibenreif (ETH Zurich) and Pascal Chang (Disney Research) will talk about self-supervised learning for satellite imagery and more! (RSVP below.)

Dan Reed gives a great opening dinner keynote at this year's Salishan meeting. Great overview ranging from #HPC and #AI to hyperscalars, low precision, #quantum, analog. All partial solutions that need to be combined into a coherent strategy.

✨What an insightful talk & discussion today at ETH AI Center on Scientific Inference with Diffusion Models by @StephanMandt (@UCIrvine), hosted by our Faculty member @vogt_je (@ETH_en) More talks, info, & registration: ai.ethz.ch/research/event…

I will be presenting 5 papers (and 1 blogpost!) at @iclr_conf this year 😱🎉 See you in Singapore!

2. Measuring Non-Adversarial Reproduction of Training Data in Large Language Models x.com/AerniMichael/s…

Congrats, your jailbreak bypassed an LLM’s safety by making it pretend to be your grandma! But did the model actually give a useful answer? In our new paper we introduce the jailbreak tax — a metric to measure the utility drop due to jailbreaks.

We present Thera🔥: The new SOTA arbitrary-scale super-resolution method with built-in anti-aliasing. Our approach introduces Neural Heat Fields, which guarantee exact Gaussian filtering at any scale, enabling continuous image reconstruction without extra computational cost.

We have a professorship opening in Responsible and Trustworthy AI in our department (tenured track assistant, associate, or full professor) in areas like Human-AI interaction, AI and System Design, Explainability, and Societal/Ethical impact and more: ifi.uzh.ch/en/department/……

Consistently in the top 10 best universities in the world for Mechanical and Process Engineering, D-MAVT secured place 6 again this year according to the QS World University Ranking. See the list by subject: ethz.ch/en/news-and-ev…