David Picard reposted
𝗦𝘂𝗿𝗳𝗹𝗼: 𝗖𝗼𝗻𝘀𝗶𝘀𝘁𝗲𝗻𝘁 𝟯𝗗 𝗦𝘂𝗿𝗳𝗮𝗰𝗲 𝗙𝗹𝗼𝘄 𝗠𝗼𝗱𝗲𝗹 𝘄𝗶𝘁𝗵 𝗚𝗹𝗼𝗯𝗮𝗹 𝗦𝘁𝗮𝘁𝗲 Antoine Guédon, Shu Nakamura, Nicolas Dufour ... Angjoo Kanazawa arxiv.org/abs/2606.13644 Trending on scholar-inbox.com
Who's Who of AI
David Picard
Professor of Computer Vision/Machine Learning at Imagine/LIGM, École nationale des Ponts et Chaussées @ecoledesponts.bsky.social Music & overall happiness 🌳🪻 Born well below 350ppm 😬 mostly silly personal views
📍Paris 🔗 https://davidpicard.github.io/
What they're sharing
Surflo: Consistent 3D Surface Flow Model with Global State arxiv.org
View on Bluesky →
arxiv.org
huggingface.co
View on Bluesky →
Grounding Functional Similarity by Invariance-Aware Model Stitching arxiv.org
View on Bluesky →
Back into Plato's Cave: Examining Cross-modal Representational Convergence at Scale arxiv.org
arxiv.org
arxiv.org
View on Bluesky →
arxiv.org
View on Bluesky →
huggingface.co
View on Bluesky →
davidpicard.github.io
killedbyapixel.github.io
View on Bluesky →
Articles & links
🎆 New paper! "Random Process Flow Matching: Generative Implicit Representations of Multivariate Random Fields", by Julien Lalanne, accepted to ICML'26 🥳 We're proposing flow-matching for inpainting in ultra-sparse setup, with applications to seismic interpolation. 📜 arxiv.org/…
David Picard reposted
🎉 Our work MIRO is accepted to #ICML2026 @icmlconf.bsky.social We integrate human preferences directly during pretraining with multi-reward conditioning. ⚡MIRO is 19x faster than baselines and 370x cheaper at inference! 🤗 Try out the models: huggingface.co/spaces/nicol... See …
David Picard reposted
Accepted paper at ICML 2026 Grounding Functional Similarity by Invariance-Aware Model Stitching The paper is available at: arxiv.org/abs/2505.20142
Ici! arxiv.org/abs/2604.18572
Vu ici aussi : arxiv.org/abs/2604.00696 Même avec un seul exemple en test-time adaptation, le RL permet d'explorer des réalisations qui généralisent par rapport à celles de train et donc probablement de renforcer l'utilisation d'un comportement de type raisonnement par rapport…
David Picard reposted
Our ICML'26 "Bayesian Two Cultures" Take on BNNs: 1/ Every 2nd paper: "MCMC for BNNs is infeasible. We thus propose [heuristic uncertainty method]." --- This is a misconception ❌
David Picard reposted
Everything is fully open-sourced, including the codebase, the model + all individual single reward model variants! 🌐 Site: nicolas-dufour.github.io/miro 📄 Paper: arxiv.org/abs/2510.25897 🛠️ Git: github.com/nicolas-dufo... 🤗 HF: huggingface.co/nicolas-dufo... 🎨 Demo: huggingfac…
David Picard reposted
Everything is fully open-sourced, including the codebase, the model + all individual single reward model variants! 🌐 Site: nicolas-dufour.github.io/miro 📄 Paper: arxiv.org/abs/2510.25897 🛠️ Git: github.com/nicolas-dufo... 🤗 HF: huggingface.co/nicolas-dufo... 🎨 Demo: huggingfac…
👏 Folks! If you are curious about the Generative Modeling via Drifting paper, but you find it difficult to understand → I wrote a different interpretation of it. It's called: "An Expectation-Maximization interpretation of Generative Modeling via Drifting" davidpicard.github.io…
David Picard reposted
Announcing the LittleJS Mini Arcade! 🚂🕹️ killedbyapixel.github.io/LittleJS-AI/... 50+ classic games built with Claude Code on my open-source HTML5 engine. Each one is a single HTML file you can fork into your own project. 😻👍 #gamedev #ClaudeCode #opensource
Their own posts
Recent commentary
Interestingly, I have no idea if the training of this generative model works or not 😅
Their network
In David Picard's orbit
Center = David Picard. Left = members they follow (green edges). Right = members who follow them (blue edges). Top = mutual follows (orange edges, slightly larger). Drag any node to reposition; click to open that profile.