KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 17/09/2026KE:SAI is an open science lab! We talk about our research without secrets. Visit our new website to find some of our recent talks: kesai.eu/talks/ 064
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 10/09/2026Making progress measurable in autonomous driving. Our new post traces the path from NAVSIM to the AlpaSim challenge: kesai.eu/blog/2026-09... 042
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 10/09/2026At Cyber Valley Day, we had the great pleasure to welcome the minister president of Baden Württemberg Cem Özdemir to our KE:SAI garage and present our mission! 031
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 08/09/2026KE:SAI is ready for CyberValley Day 2026! cyber-valley.de/en/events/cy... 052
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 23/07/2026KE:SAI is hiring! You are super excited about world models and autonomy? Come and join our team! Let's have impact together! kesai.eu/join/ 0164
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 17/07/2026It's getting real! We have been working on self-driving for more than 15 years, and most of that work has happened in simulation. Today that changes: our research vehicle arrived, and we can start testing our ideas in the physical world. kesai.eu/blog/2026-07... 0216
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 10/07/2026In this talk from the GenAV Workshop at ICRA 2026, our CTO Kashyap Chitta highlights how open tools, shared benchmarks, and collaborative research can help accelerate progress in autonomous driving. Watch here: www.youtube.com/watch?v=_CYp...youtube.comKeshyap Chitta - KE:SAI - Workshop on Generalization in Autonomous Driving at ICRA 2026What does it take to make autonomous driving research more open, accessible, and scalable? In this talk from the 1st GenAV Workshop at ICRA 2026, Kashyap Chitta explores Democratizing Autonomous… 072
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 08/07/2026Hosted by NVIDIA's Autonomous Vehicle Research Group, KE:SAI, and HKU. Start/register: huggingface.co/spaces/nvidi... Questions/discussions: github.com/NVlabs/alpasimhuggingface.coAlpaSim Challenge 2026 - a Hugging Face Space by nvidiaUsers can browse all details of the AlpaSim End‑to‑End Closed Loop Challenge—including overview, tracks, timeline, prizes, and documentation—directly in their browser. An embedded Gradio interface ... 000
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 08/07/2026Key dates: Aug 15: mid-competition checkpoint Sep 15: rules/submission format freeze Oct 31: public leaderboard closes + technical reports due Nov 15: final results + award decisions Each track awards two NVIDIA DGX Spark prizes. 100
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 08/07/2026Leaderboards report an ELO-style Policy Capability Score and Avg Distance Between At-Fault Incidents. Driver instances are capped at 16 GiB VRAM; policies should target <=0.1 s/model forward pass for a 10 Hz loop, replicating realistic deployment constraints. 100
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 08/07/2026Participants get public development data, starter tools, documentation, a submission CLI, and container requirements before organizer-side evaluation on private curated scenarios. Teams submit AlpaSim-compatible driving policies, not downloadable test-set predictions. 100
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 08/07/2026Two tracks: nuPlan: public-resource track for nuPlan/NAVSIM workflows; award entries may use only public datasets, annotations, and model weights. PAI-AV: larger-scale track; private resources allowed for awards with usage rights and high-level disclosure. 100
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 08/07/2026Why closed loop? Open-loop scores against logged trajectories miss compounding errors. In AlpaSim, submitted policies control the ego vehicle over time, so small planning mistakes can change the future scene distribution and expose failures that static prediction scores miss. 100
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 08/07/2026The AlpaSim E2E Closed Loop Challenge 2026 is open for registration and leaderboard access in a soft-open period. Build AV policies for realistic closed-loop simulation, where each policy's own decisions shape future scenes, observations, and interactions. 182
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 04/07/2026🔥 ICML 2026 paper alert! Introducing PointDiT, a radically simple diffusion transformer for 3D geometry generation that bypasses the VAE bottleneck completely. We asked: Can we make 3D reconstruction beautifully simple? Turns out: Yes! 📄 Paper + Demo: haofeixu.github.io/pointdit/ 092
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 03/07/2026Check out the video of our approach and the leaderboard below: Video: www.youtube.com/watch?v=hsck... Leaderboard: kitscenes.com/benchmarks/l...youtube.comKITScenes LongTail 3rd Place 000
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 03/07/2026Our team placed 3rd in the KITScenes LongTail Challenge at the CVPR 2026 Workshop on Autonomous Driving! We used zero-shot evaluation with the NVIDIA Alpamayo 1.5 vision-language-action model, with zero fine-tuning or task-specific training. 1103
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 03/07/2026Our CTO Kashyap gave a very nice talk at ICRA 2026 on "World Models: The Next Frontier of Motion Prediction". The talk sketches the history and also lays out what we plan to do at KE:SAI. Watch here: www.youtube.com/watch?v=sxlh...youtube.comKashyap Chitta: World Models: The Next Frontier of Motion PredictionTalk given on the 8th Workshop on Long-term Human Motion Prediction (LHMP) at ICRA 2026. Link to the Workshop website: https://motionpredictionicra2026.github.io Talk Abstract: Predicting how the… 063
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 28/06/2026World Engine is a RL post-training simulator for end-to-end driving. In collaboration with industry (and industry-scale data), we present an end-to-end policy post-trained within World Engine that achieves 200 km real driving without intervention. github.com/OpenDriveLab... 062
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 27/06/2026Link to blog: kesai.eu/blog/2026-06... 020
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 27/06/2026The unreasonable effectiveness of state-of-the-art imitation learning models in the CARLA simulator is largely explained by a hidden shortcut: the target point bias. Read more in our new blog. Link in the comments. 170
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 24/05/2026KE:SAI has been (briefly) featured in ZDF Länderspiegel! Watch here (starting at 17:00, in German): www.zdf.de/video/magazi... 094
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 20/05/2026Bill Gurley made a fantastic post on the case for open-science labs like KE:SAI. A strongly encouraged read on open source in general and for autonomous vehicles in particular! p3institute.substack.com/p/from-open-...p3institute.substack.comFrom Open Source Software to Open Source StrategyHow the Smartest Executives Are Using Open Source Techniques to Optimize Corporate Strategy 061
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 20/05/2026KE:SAI is hiring! Join our mission at: kesai.eu/join/kesai.euKE:SAI — Open Positions at Our AI Research LabJoin KE:SAI, a Franco-German non-profit AI research lab co-located in Tübingen and Paris. Open positions for researchers in physical AI and autonomous systems. 010
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 20/05/2026The research is led by a founding team of experts: Andreas Geiger (CEO), Kashyap Chitta (CTO), and Bernhard Schölkopf, alongside co-founders Bernhard Jaeger and Daniel Dauner. 110
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 20/05/2026KE:SAI has offices in Tübingen and Paris. By integrating foundational open research and the agility of modern startups, KE:SAI will develop a fully open technology stack for physical AI. 100
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 20/05/2026@kyutai-labs.bsky.social and @ellisinsttue.bsky.social announce the official launch of KE:SAI – Kyutai ELLIS Scalable Autonomous Intelligence. KE:SAI is a non-profit open science research lab dedicated to the next frontier of AI: systems that can understand and act in the world. 100
KE:SAI - Kyutai ELLIS Scalable Autonomous Intelligence @kesai.eu · 20/05/2026Today @kyutai-labs.bsky.social and @ellisinsttue.bsky.social launch @kesai.eu! Robot learning is bottlenecked by the cost of physical interaction. Our mission is to advance the efficiency frontier of robust & safe physical AI through fully open and reproducible research. kesai.eu/blog/2026-05...kesai.euKyutai and ELLIS Tübingen launch KE:SAI: A Premier Franco-German Partnership for Open Science in Physical AIKyutai and ELLIS Tübingen announce the official launch of KE:SAI – Kyutai ELLIS Scalable Autonomous Intelligence, a non-profit open science lab for physical AI. 11710