ADTool: open-source python library facilitating the use of curiosity-driven exploration to discover diverse behaviors in a complex system, with many examples and nice visualisation tools! 🔍 Great to see this out, congrats to the team at Inria Flowers team👇👏

M.I.A. Seconde "Lancement de l’outil numérique #MIASeconde pour la remédiation en français et mathématiques" site.ac-martinique.fr/drane/?p=5130 Etude d'impact pour mesurer l'aide de l'#IA contre le décrochage #TestdePositionnementSeconde #DEPP EvidenceB LIP6 Inria Flowers team Pierre-Yves Oudeyer

🚨New preprint🚨 When testing LLMs with questions, how can we know they did not see the answer in their training? In this new paper we propose a simple out of the box and fast method to spot contamination on short texts with Stefano Palminteri (@stepalminteri.bsky.social) and Pierre-Yves Oudeyer ! 1/16

I'm excited to announce that this work has been accepted at NeurIPS Conference 🥳 We hope to spark cross-disciplinary conversations on goal selection in biological and artificial agents. Check it out at openreview.net/forum?id=GbqzN… Cédric Pierre-Yves Oudeyer Anne Collins Inria Flowers team Collins Lab

Large Language Models #LLMs are black boxes. In this thread, I am going to summarize a couple of (open) tools that we developed (mainly the amazing Nicolas Yax with the invaluable help of Pierre-Yves Oudeyer ) to solve some of these issues and empower robust and reproducible research

« L’IA générative va contribuer à modeler nos relations sociales et notre relation au savoir » : « il y a urgence à se former. » Une passionnante table ronde à #EducatechExpo avec Axel Jean Margarida ROMERO #creacube 🐈🐸🍁 Jean Cattan … cafepedagogique.net/2024/11/15/il-…

#IA et #Éducation au présent, enjeux et actions en France et en Europe Éducation Numérique Ministère Éducation nationale à Educatech Expo 2024 Ici des éléments par le Le café pédagogique /c Audran Le Baron Pierre-Yves Oudeyer Didier Roy droy.bsky.social @elieallouche Clément Fantoli Alain Thillay Chaire Unesco RELIA AI4T #CREIA #GTNum

[Podcast📻] Comment appréhender l' #IA à l'école ? Peut-on modéliser les mécanismes d'apprentissage et de curiosité pour mieux comprendre le plaisir d'apprendre ? 🔎Retrouvez l'interview de Pierre-Yves Oudeyer, chercheur @inria par Jerome Colombain | Monde Numérique autour des travaux de Inria Flowers team. ⤵️

🚀 Exciting Internship Opportunities for AI and CogSci Students🌟 Join Inria Flowers team and work on these cool topics: 🔧 Curriculum learning of skill libraries in autotelic agents using LLMs and program synthesis wth Pourcel Julien 🎯 Balancing Exploration and Exploitation in

Stop by NeurIPS Conference poster #3909 11 am - 2 pm today if you want to wish me a happy birthday, hear about goal selection in RL, or both! Cédric Anne Collins Pierre-Yves Oudeyer Collins Lab Inria Flowers team

1/ If you want to know how to generate diverse and challenging Programming Puzzles stop by our NeurIPS Conference poster #3010 (East Exhibit Hall A-C, 4:30-7:30 pm) Cédric Gaia Molinaro 🦋 gaiamolinaro.bsky.social Pierre-Yves Oudeyer Laetitia Teodorescu

Carta Thomas and Loris Gaven will be presenting SAC-GLAM at IMOL Workshop | NeurIPS 2024 tomorrow. #NeurIPS2024 TLDR: We traded PPO for SAC in GLAM. We get better sample efficiency for the same compute budget. But more importantly, we can now use HER and get even more sample efficient!

#AIActionSummit | Comment l’IA participe aux apprentissages ? Quels impacts sur l'#education ? Dans le cadre des Samedis Sciences avec #SAPS Université de Bordeaux, rencontrez Hélène Sauzéon (Inria Flowers team) : 🗓️le 8/02 à 16h30 📍Médiathèque #Gradignan ➕infos : urlr.me/VZ2YfR

💡 L'intelligence artificielle transforme tous les domaines… mais sommes-nous prêts ? Alors que le Sommet pour l’Action sur l’IA se tient à Paris, réunissant dirigeants, experts et entreprises du monde entier, une question cruciale se pose : comment permettre à chacun de

🚀 Introducing 🧭MAGELLAN—our new metacognitive framework for LLM agents! It predicts its own learning progress (LP) in vast natural language goal spaces, enabling efficient exploration of complex domains.🌍✨Learn more: 🔗 arxiv.org/abs/2502.07709 #OpenEndedLearning #LLM #RL

Max Taylor-Davies will present our new paper at the Evostar on Thu. Emergent Kin Selection of Altruistic Feeding via Non-Episodic Neuroevolution arxiv.org/abs/2411.10536 With Gautier Hamon and TimotheBoulet , leveraging previous work with Eleni Nisioti Summary:

Hundreds of new generative AI models, some close, some open, are released every months. Most of them derive in some way (weights, data, architecture,etc) from previous models, but this information is rarely disclosed. Can one infer their ancestors? Yes, with PhyloLM 👇

Rohan Pandey . Rohan Pandey here are several papers on curiosity-driven LLM agents: - MAGELLAN (ICML 2025): arxiv.org/abs/2502.07709 - ACES (Neurips 2024): proceedings.neurips.cc/paper_files/pa… - LMA3: proceedings.mlr.press/v232/colas23a.… Cédric - Perspective and history (before 2010s :)): nature.com/articles/s4225…