AI for Global Goals | OxML

From ideas to impact—our participants reflect on their MLx experience and what they’re taking forward. Register now to be part of the next cohort. oxfordml.school/2026 #OxML #MachineLearning #AI #GenerativeAI #RepresentationLearning #AICommunity #DeepLearning #MLResearch #AIEducation #FutureOfAI #LearnAI #AIEvents #OxfordAI

Explore how AI systems can perceive, reason, and act in the real world. Join Oier Mees (Microsoft, ETH Zürich) for a deep dive into Embodied Multimodal Intelligence with Foundation Models, where vision, language, and action come together to shape the next generation of intelligent agents. MLx Representation Learning & Generative AI University of Oxford Mathematics Discover how foundation models are moving beyond text and images into real-world interaction. Secure your place: oxfordml.school/2026 In-person seats are limited. Register by 22 May 2026. #OxML #MultimodalAI #RepresentationLearning #FoundationModels #Robotics @oier_mees @Microsoft @ETH_en @OxUniMaths

A major frontier in AI is no longer just making models larger. It is building systems that can perceive, reason, and act in the real world. That is where embodied multimodal intelligence comes in. As foundation models move beyond static benchmarks, one of the biggest questions is how to turn them into agents that can operate across modalities, interact with physical environments, and generalise across tasks. This is where robotics, multimodal learning, and reasoning begin to converge. At #OxML 2026, Oier Mees (@Microsoft, ETH Zurich) will explore embodied multimodal intelligence with foundation models, including robot foundation models, generalist robot policies, and reasoning-enhanced robotics. This is exactly the kind of topic that matters for those who want to work on where modern AI is heading next. 👉 oxfordml.school/2026 #OxML2026 #MachineLearning #AI #MultimodalAI #FoundationModels #Robotics #DeepLearning, @oier_mees, @ETH_en

One of the most important frontiers in AI is turning foundation models into systems that can perceive, reason & act in the real world. At #OxML 2026, we'll explore embodied multimodal intelligence, generalist robot policies & reasoning-enhanced robotics. oxfordml.school

📣 MLx Representation Learning & Generative AI lecture update: 📍On Causal Discovery and the Extrapolation of Causal Effects by Prof Ricardo Silva from UCL 🗓 15–18 July | Mathematical Institute, Oxford 🌐 oxfordml.school Revisit how causal effect learning predicts how the world responds to an agent’s actions using off-policy data, while causal discovery seeks to uncover the underlying structure that makes such extrapolation possible. The lecture also explores how outcomes from past actions across different environments can be used to calibrate causal discovery, making these predictions more reliable. Registrations are now open! #OxML2026 #CausalEffects #RepLearning #AIAgent

@monalinejad @ucl A timely and important topic for anyone working on reliable decision-making, interventions, and generalisation beyond observed data. Excited to have Prof. Silva speaking at OxML 2026.

Causal inference becomes much harder when the structure is uncertain. Estimating effects from off-policy data is not enough on its own. At #OxML 2026, Prof. Ricardo Silva (@ucl) will explore causal discovery and the extrapolation of causal effects. 👉oxfordml.school

📣 Announcing the OxML 2026 Poster Challenge We are excited to launch the OxML 2026 Poster Challenge, taking place this July at the Mathematical Institute, University of Oxford. The challenge highlights outstanding research and applied work across machine learning and AI, and offers a platform to present your work, receive feedback, and engage with the OxML community. Submissions are invited in two tracks: 🟢 MLx Health & Bio 🟢 MLx Representation Learning & Generative AI 🏆 Top posters in each track will receive $1,000 awards. 📅 Submission deadline: 15 May 2026 We particularly encourage submissions from researchers preparing or submitting work to venues such as NeurIPS, ICML, ICLR, and ACL, as well as early-stage and applied contributions. Submit via EasyChair: easychair.org/conferences/?c… Full call for posters: easychair.org/cfp/oxml26 #OxML2026 #OxML #MachineLearning #AI #GenerativeAI #RepresentationLearning #HealthAI #BioAI, @GRESEARCHjobs

Multimodal AI will shape the next phase of ML. As AI moves beyond single-modality systems, stronger foundations in fusion, alignment, reasoning, & generation will matter even more. At #OxML 2026, we’ll explore the principles behind multimodal AI. 👉oxfordml.school/2026

Looking back at OxML 2025 – MLx Representation Learning & Generative AI. A week filled with inspiring lectures, engaging discussions, and meaningful connections. With participants joining from 100+ countries, it was amazing to see such a global community come together to learn, share ideas, and explore the future of representation learning and generative AI. We’re excited to continue building this community at OxML 2026. Registrations are open: oxfordml.school 2025 Speakers featured: Ashley Edwards, Edward Johns,Christian Rupprecht, Aymeric Dieuleveut, Gerhard Neumann, Yingzhen Li, Xuan-Son Nguyen, Fazl Barez, Mergen Nachin, Abdul Fatir Ansari 2025 Committee: Mona Alinejad, D.Phil. (Oxon), Reza Khorshidi, D.Phil. (Oxon), Yali Du, Karo Moilanen #OxML2026 #OxfordMachineLearning #MachineLearning #RepresentationLearning #GenerativeAI #DeepLearning #AICommunity

Oxford Machine Learning School (OxML) returns for its 7th edition in 2026, bringing together leading researchers and practitioners in artificial intelligence and machine learning. The programme offers advanced, research-driven courses taught by world-class AI scientists and designed for PhD students, researchers, and industry practitioners who want to deepen their expertise in modern machine learning. Participants can choose from three specialised tracks: • MLx Cases – Hands-on development of generative AI systems and real-world applications • MLx Health & Bio – AI for healthcare, biomedicine, and life sciences • MLx Representation Learning & Generative AI – Foundation models, modern machine learning, and AI for mathematics Courses combine theoretical foundations with practical system-level insights, providing a rigorous learning experience at the intersection of research and applied AI. 📍 Oxford, UK & Online | May & July 2026 🔗 oxfordml.school #OxML #OxML2026 #MachineLearning #AI #DeepLearning #GenerativeAI @CIFAR_News @oxmartinschool @OxUniMaths

We’re excited to share the speaker lineup for MLx Representation Learning & Generative AI at OxML 2026. MLx Representation Learning & Generative AI explores the latest advances in representation learning and the rapidly evolving landscape of generative AI and frontier models. The program also features the OxML 2026 Poster Challenge, a competitive research poster session where researchers, students, and practitioners present and discuss new work across machine learning and AI. 📍 Oxford Mathematical Institute & Online 📅 15–18 July 2026 📝 Register by May 22. oxfordml.school/register Current speakers include: • Kevin Buzzard — Imperial College London • Tom Rainforth — University of Oxford • Amanda Prorok — University of Cambridge • Oier Mees — Microsoft, ETH Zürich • Petar Veličković — Google DeepMind • Tim Rocktäschel — UCL • Ricardo Silva — UCL • Paul Liang — Massachusetts Institute of Technology • Fazl Barez — University of Oxford • Alexander Tong— Aithyra • Peter Gostev — Arena AI Join us in Oxford to learn, exchange ideas, and connect with the global ML community. #OxML #MachineLearning #GenerativeAI #RepresentationLearning #AIResearch #OxfordML @UniofOxford @Cambridge_Uni @MIT @MSFTResearch @ucl @arena @aprorok @FazlBarez @_rockt @tom_rainforth @AlexanderTong7 @pliang279 @PetarV_93 @oier_mees

MLx Cases Practical Lecture Announcement We’re excited to welcome Peter Gostev, AI Capability Lead at Arena AI, for an upcoming MLx Cases Practical Lecture. Topic: Hands-On with Today’s Cutting-Edge Tools: Building Practical Apps from a Real-World Example In this session, Peter will walk participants through a hands-on exploration of modern AI development tools, demonstrating how to turn a real-world use case into a practical working application. The lecture will focus on the current generation of tools (such as code-generation and agentic AI systems) and how they can accelerate the process from idea to prototype. 🚀 MLx Cases is designed for practitioners who want to design, build, and own AI systems — a durable, career-long advantage. Register here: oxfordml.school/2026 #MLxCases #GenerativeAI #AppliedAI #OxML2026 Yali Du, OxML @petergostev @arena

🚀 New MLx Case Announced: Generative AI for Calendar Booking & Travel Planning Featuring a practical session by Xin Eric Wang (University of California, Santa Barbara) We explore how Generative AI can power real-world agentic systems — from intelligent calendar coordination to automated travel planning. This session goes beyond theory. We’ll look at how modern foundation models can: 💠Understand scheduling constraints across multiple users 💠Plan and optimise travel itineraries 💠Integrate with real-world tools and APIs 💠Handle ambiguity, preferences, and changing conditions Perfect for practitioners who want to design and deploy AI systems that actually do things — not just generate text. 🚀 MLx Cases is designed for practitioners who want to design, build, and own AI systems — a durable, career-long advantage. Register here: oxfordml.school/2026 #MLxCases #GenerativeAI #AIAgents #AppliedAI UC Santa Barbara, Yali Du. OxML

We’re excited to welcome Noor Sajid, Kempner Research Fellow at Harvard University and Guest Scientist at the Max Planck Institute for Biological Cybernetics, whose research bridges machine learning and computational neuroscience. Featuring a lecture on Building a CNN for Image Classification at MLx Cases. From convolutional layers and feature maps to training pipelines and evaluation, this session will walk through how to design and implement a Convolutional Neural Network for real-world image tasks — grounding theory in practical system-building. 🚀 MLx Cases is designed for practitioners who want to design, build, and own AI systems — a durable, career-long advantage. Register here: oxfordml.school/2026 #MLxCases #OxML #OxML2026 #OxML2026 #ConvolutionalNeuralNetwork Noor S., Yali Du, OxML

🎥 What is the OxML experience like? Hear it from our participants. our participants share how OxML was especially valuable for those getting into or just beginning their careers in machine learning and AI. From building foundational understanding to gaining exposure to real research and practical systems, OxML creates an environment where early-career participants can ask questions, engage with experts, and see how ideas translate into impact. If you're starting your journey in ML, this is the kind of community and clarity that makes a difference. 🚀 Join us at OxML 2026. oxfordml.school #OxML #OxML2026 #MachineLearning #AICommunity #Oxfordsummerschool #AppliedML

🎓 Featuring a lecture on Building Diffusion Models for Images & Sequences Diffusion models have reshaped generative AI — from image synthesis to sequence generation. In this MLx Cases session, we go beyond theory to unpack how diffusion models are actually built: architecture design, training dynamics, scaling considerations, and how to adapt them from images to structured sequences. This is a practical, system-level deep dive into what it takes to implement and deploy diffusion models in real-world settings — not just use them, but truly understand and engineer them. The lecture will be delivered by Jialin Yu, who brings strong expertise at the intersection of generative modelling and applied AI systems, bridging research insight with hands-on implementation. 🚀 MLx Cases is designed for practitioners who want to design, build, and own AI systems — a durable, career-long advantage. Register here: oxfordml.school/2026 #MLxCases #OxML #OxML2026 #diffusionmodels #aiimaging #generativeai

Education should never be interrupted by repression. Today, AI for Global Goals launches the Crisis Access Scholarship – Iran for the OxML 2026 cohort. The scholarship provides a full waiver for the online components of the programme to exceptional applicants whose education has been disrupted by recent violence and repression in Iran. Founded by an Iranian academic, this initiative is deeply personal. Talent should not be silenced, and access to education should not depend on circumstance. To apply, visit: forms.gle/QSG3a1PVjcRBt8… #OxML #OxML2026 #MachineLearning #AIForGood #AccessToEducation #Iran

Build your own GPT — from scratch. 🧠⚙️ In this hands-on MLx Cases lecture, Sizhe Yuen from The Alan Turing Institute will walk you through: Building a MiniGPT This session is ideal for practitioners who want to deeply understand how GPT-style models work under the hood and gain the confidence to build and adapt them for real-world applications. 🚀 MLx Cases is designed for practitioners who want to design, build, and own AI systems — a durable, career-long advantage. Register here: oxfordml.school/2026 #MLxCases #OxML #OxML2026 #MiniGPT #LLMs