BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//MBZUAI - ECPv6.15.16.1//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:MBZUAI
X-ORIGINAL-URL:https://mbzuai.ac.ae
X-WR-CALDESC:Events for MBZUAI
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:Europe/Paris
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20240331T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20241027T010000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20250330T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20251026T010000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:20260329T010000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:20261025T010000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=Europe/Paris:20250212T090000
DTEND;TZID=Europe/Paris:20250213T161000
DTSTAMP:20260526T222720
CREATED:20250130T083524Z
LAST-MODIFIED:20250223T123750Z
UID:20917-1739350800-1739463000@mbzuai.ac.ae
SUMMARY:MBZUAI Workshop 2025
DESCRIPTION:Home\nSpeaker List\nProgram Day 1\nProgram Day 2\n\n\nWorkshop Goals\nWe are pleased to announce the MBZUAI Workshop 2025 on “Foundations and Advances in Generative AI”: Theory and Methods\, organized by the Machine Learning Department of Mohamed bin Zayed University of Artificial Intelligence (MBZUAI) in collaboration with the MBZUAI France Lab. Following the first event successfully held in Abu Dhabi\, this workshop aims to foster collaboration and accelerate progress in the machine learning aspects of large language models (LLMs)\, multimodal model research\, and foundations of Generative AI. \nScheduled for February 12–13\, the program includes: \n\nInvited Talks: Two days of presentations by leading experts on cutting-edge machine learning and generative AI advancements.\nPoster Session: A platform for researchers to showcase their work.\nJoin us for insightful discussions and networking opportunities in this rapidly evolving field!\n\n\n\nRegistration\nTo participate in the workshop\, please complete the registration form. If you intend to present a poster\, kindly ensure you fill out the appropriate fields before February 3\, 2025\, so we can review it. \n\n\n\nAreas of Focus\nWe invite contributions on topics broadly related to machine learning for large models. Key areas of focus will include: \n\nDevelopment of novel ML architectures for large model training and inference.\nAddressing bias and fairness in large model training data and output.\nTechniques for interpretability and explainability of large model behavior.\nRepresentation learning for large multimodal models.\nMitigating risks and ensuring safety in large model development.\nScaling and efficiency considerations for large-scale model training.\nApplication domains including bio/medical and others.\n\n\n\nOrganizers\nOrganizing committee: \n\nEric Moulines\, Ecole Polytechnique & MBZUAI\nGuokan Shang\, MBZUAI France Lab\nMichalis Vazirgiannis\, Ecole Polytechnique & MBZUAI\nKun Zhang\, MBZUAI\n\nLogistics support: \n\nGeorgia Dimopoulos\, MBZUAI France Lab\nBrenda Ward\, MBZUAI\n\n\n Directions \n\n    \n    \n\nGitHub Page \n\n\n\nInvited Speakers\n\n\n    \n    \n      Gérard Biau\n      Sorbonne University\n    \n  \n\n    \n    \n      Emmanuel Candès\n      Stanford University\n    \n  \n\n   \n    \n      Valentin De Bortoli\n      Google DeepMind (on leave CNRS)\n    \n  \n\n\n\n    \n    \n      Alexandre Défossez\n      Kyutai\n    \n  \n\n    \n    \n      Yazid Janati\n      École Polytechnique\n    \n  \n\n    \n    \n      Lingpeng Kong\n      University of Hong Kong\n    \n  \n\n\n\n    \n    \n      Salem Lahlou\n      MBZUAI\n    \n  \n\n    \n    \n      Preslav Nakov\n      MBZUAI\n    \n  \n\n   \n    \n      Giannis Nikolentzos\n      University of Peloponnese\n    \n   \n\n\n\n    \n    \n      Maks Ovsjanikov\n      Google DeepMind & École Polytechnique\n    \n  \n\n    \n    \n      Thomas Pierrot\n      InstaDeep\n    \n  \n\n    \n    \n      Gaël Richard\n      Télécom Paris\n    \n  \n\n\n\n    \n    \n      Martin Takáč\n      MBZUAI\n    \n  \n\n    \n    \n      Jie Tang\n      Tsinghua University\n    \n  \n\n   \n    \n      Daniil Tiapkin\n      École Polytechnique\n    \n  \n\n\n\n    \n    \n      Michal Valko\n      INRIA & Stealth Startup\n    \n  \n\n    \n    \n      Julien Velcin\n      University of Lyon\n    \n  \n\n    \n    \n      Eric Xing\n      MBZUAI & Carnegie Mellon University\n    \n  \n\n\n\n    \n    \n      Kun Zhang\n      MBZUAI\n    \n  \n\n    \n    \n  \n\n \n\n\nProgram on Wednesday\, February 12\n\n\n9:00 AM\nRegistration and Coffee & Tea!\n9:30 AM\nOpening RemarksEmmanuel Candès (Stanford University) \nWe present new statistical methods for obtaining validity guarantees on the output of large language models (LLMs). These methods enhance conformal prediction techniques to filter out claims/remove hallucinations while providing a finite-sample guarantee on the error rate of what it being presented to the user. This error rate is adaptive in the sense that it depends on the prompt to preserve the utility of the output by not removing too many claims. We demonstrate performance on real-world examples. This is joint work with John Cherian and Isaac Gibbs. \n10:50 AM\nCoffee & Tea Break\n11:00 AM\nThe ChatGLM’s Road to AGIGaël Richard (Télécom Paris) \nWe will describe and illustrate the concept of hybrid (or model-based) deep learning for music generation. This paradigm refers here to models that associates data-driven and model-based approaches in a joint framework by integrating our prior knowledge about the data in more controllable deep models. In the music domain\, prior knowledge can relate for instance to the production or propagation of sound (using an acoustic or physical model) or how music is composed or structured (using a musicological model). In this presentation\, we will first illustrate the concept and potential of such model-based deep learning approaches and then describe in more details its application to unsupervised music separation with source production models\, music timbre transfer with diffusion and symbolic music generation with transformers using structured informed positional encoding. \n12:20 PM\nAuditing and Mitigating Biases in (compressed) Language ModelsMichal Valko (INRIA & Stealth Startup) \nEnsuring alignment of language models’ outputs with human preferences is critical to guarantee a useful\, safe\, and pleasant user experience. Thus\, human alignment has been extensively studied recently and several methods such as Reinforcement Learning from Human Feedback (RLHF)\, Direct Policy Optimisation (DPO) and Sequence Likelihood Calibration (SLiC) have emerged. In this paper\, our contribution is two-fold. First\, we show the equivalence between two recent alignment methods\, namely Identity Policy Optimisation (IPO) and Nash Mirror Descent (Nash-MD). Second\, we introduce a generalisation of IPO\, named IPO-MD\, that leverages the regularised sampling approach proposed by Nash-MD. @This equivalence may seem surprising at first sight\, since IPO is an offline method whereas Nash-MD is an online method using a preference model. However\, this equivalence can be proven when we consider the online version of IPO\, that is when both generations are sampled by the online policy and annotated by a trained preference model. Optimising the IPO loss with such a stream of data becomes equivalent to finding the Nash equilibrium of the preference model through self-play. Building on this equivalence\, we introduce the IPO-MD algorithm that generates data with a mixture policy (between the online and reference policy) similarly as the general Nash-MD algorithm. We compare online-IPO and IPO-MD to different online versions of existing losses on preference data such as DPO and SLiC on a summarisation task. \n14:40 PM\nMoshi: A Speech-text Foundation Model for Real-time DialogueMaks Ovsjanikov (Google DeepMind & École Polytechnique) \nRecent works have shown that\, when trained at scale\, uni-modal 2D vision and text encoders converge to learned features that share remarkable structural properties\, despite arising from different representations. However\, the role of 3D encoders with respect to other modalities remains unexplored. Furthermore\, existing 3D foundation models that leverage large datasets are typically trained with explicit alignment objectives with respect to frozen encoders from other representations. In this talk I will discuss some results on the alignment of representations obtained from uni-modal 3D encoders compared to text-based feature spaces. Specifically\, I will show that it is possible to extract subspaces of the learned feature spaces that have common structure between geometry and text. This alignment also leads to improvement in downstream tasks\, such as zero shot retrieval. Overall\, this work helps to highlight both the shared and unique properties of 3D data compared to other representations. \n16:10 PM\nRedefining AI Reasoning: From Self-Guided Exploration to Causal Loops\, and Transformer-GNN FusionValentin De Bortoli (Google DeepMind (on leave CNRS)) \nDiffusion models have revolutionized generative modeling. Conceptually\, these methods define a transport mechanism from a noise distribution to a data distribution. Recent advancements have extended this framework to define transport maps between arbitrary distributions\, significantly expanding the potential for unpaired data translation. However\, existing methods often fail to approximate optimal transport maps\, which are theoretically known to possess advantageous properties. In this talk\, we will show how one can modify current methodologies to compute Schrödinger bridges—an entropy-regularized variant of dynamic optimal transport. We will demonstrate this methodology on a variety of unpaired data translation tasks. \n10:10 AM\nMulti-modal Foundation Models for BiologyGiannis Nikolentzos (University of Peloponnese) \nGraph generation has emerged as a crucial task in machine learning\, with significant challenges in generating graphs that accurately reflect specific properties. In this talk\, I will present Neural Graph Generator\, our recently released model which utilizes conditioned latent diffusion models for graph generation. The model employs a variational graph autoencoder for graph compression and a diffusion process in the latent vector space\, guided by vectors summarizing graph statistics. Overall\, this work represents a shift in graph generation methodologies\, offering a more practical and efficient solution for generating diverse graphs with specific characteristics. \n11:40 AM\nA Primer on Physics-informed Machine LearningSalem Lahlou (MBZUAI) \nGenerative Flow Networks offer a framework for sampling from reward-proportional distributions in combinatorial and continuous spaces. They provide an alternative to established methods such as MCMC that suffer from slow mixing in high-dimensional spaces. By leveraging flow conservation principles\, GFlowNets enable exploration in scenarios where the diversity of solutions is crucial\, differing from traditional reinforcement learning and generative models. The framework has shown practical utility in molecular design\, protein structure prediction\, and Bayesian network discovery\, particularly when dealing with noisy reward landscapes where maintaining sample diversity is essential. Recent works have also explored GFlowNets as a mechanism for improving the systematic exploration capabilities of large language models. This talk will present the theoretical foundations of GFlowNets and discuss current research directions in expanding their applications. \n13:00 PM\nLunch\n14:00 PM\nWhat’s not an Autoregressive LLM?Kun Zhang (MBZUAI) \nCausality is a fundamental notion in science\, engineering\, and even in machine learning. Uncovering the causal process behind observed data can naturally help answer ‘why’ and ‘how’ questions\, inform optimal decisions\, and achieve adaptive prediction. In many scenarios\, observed variables (such as image pixels and questionnaire results) are often reflections of the underlying causal variables rather than being causal variables themselves. Causal representation learning aims to reveal the underlying hidden causal variables and their relations. In this talk\, we show how the modularity property of causal systems makes it possible to recover the underlying causal representations from observational data with identifiability guarantees: under appropriate assumptions\, the learned representations are consistent with the underlying causal process. We demonstrate how identifiable causal representation learning can naturally benefit generative AI\, with image generation\, image editing\, and text generation as particular examples. \n15:20 PM\nCoffee & Tea Break\n15:30 PM\nFactuality Challenges in the Era of Large Language Models: Can we Keep LLMs Safe and Factual?Yazid Janati (École Polytechnique) \nDiffusion models have recently shown considerable potential in solving Bayesian inverse problems when used as priors. However\, sampling from the resulting denoising posterior distributions remains a challenge as it involves intractable terms. To tackle this issue\, state-of-the-art approaches formulate the problem as that of sampling from a surrogate diffusion model targeting the posterior and decompose its scores into two terms: the prior score and an intractable guidance term. While the former is replaced by the pre-trained score of the considered diffusion model\, the guidance term has to be estimated. In this paper\, we propose a novel approach that utilises a decomposition of the transitions which\, in contrast to previous methods\, allows a trade-off between the complexity of the intractable guidance term and that of the prior transitions. We also show how the proposed algorithm can be extended to handle the sampling of arbitrary unnormalised densities. We validate the proposed approach through extensive experiments on linear and nonlinear inverse problems\, including challenging cases with latent diffusion models as priors. \n16:10 PM\nDemonstration-Regularized RL and RLHF
URL:https://mbzuai.ac.ae/event/mbzuai-workshop-2025/
LOCATION:Fondation François Sommer\, 62 Rue des Archives\, France\, Paris\, France
ATTACH;FMTTYPE=image/jpeg:https://staticcdn.mbzuai.ac.ae/mbzuaiwpprd01/2025/01/mbzuai-workshop-2025-scaled.jpg
END:VEVENT
END:VCALENDAR