Pour Demain’s Post

Excited to share some recent research and policy recommendations we've been working on at Pour Demain, relating to serious incident reporting for general purpose AI models. ✒️ This paper covers: 1. Reporting timelines ⌛ 2. Unique incident reporting concerns posed by GPAI models ☣️ 3. Contents of a final incident report 📄 4. Additional recommendations ➕ Learning from our mistakes will be pivotal in building the future we want.

✨ Welcome to the second edition of MedEngine webinars! Dive into the world of Artificial Intelligence and Machine Learning with MedEngine's Lead Data Scientist, Juha Mehtälä. Explore how to harness these technologies in the context of real-world data, discover potential pitfalls, and learn when an AI approach is optimal or when it might be wise to consider alternative solutions. Join us live or sign up for the replay. Read more and register below! https://lnkd.in/d-aQXsF7 #MedEngine #Webinar #TheValueIsInTheLinks #PharmaceuticalIndustry #AI #MachineLearning

Convergence 2024 is on the horizon and we're looking for speakers! This May 8-9, the Convergence Conference will focus on the development of scalable #ML platforms, #AIGovernance, and the development of #LLMs, #FederatedML, and Vision Transformers. The conference will feature expert-led talks, panels, and workshops designed to address both the technical nuances and the broader implications of #AI regulation for enterprises. Applications are now open for speakers: #MachineLearning #GenAI

☀ Convergence 2024 is on the horizon and we're looking for speakers! 🔥 This May 8-9, the Convergence Conference will focus on the development of scalable #ML platforms, #AIGovernance, #LLMs, #FederatedML, and Vision Transformers. 👩🏫 The conference will feature expert-led talks, panels, and workshops designed to address both the technical nuances and the broader implications of #AI regulation for enterprises. 🎤 Applications are now open for speakers: #MachineLearning #GenAI Comet

Convergence 2024 is on the horizon and we're looking for speakers! This May 8-9, the Convergence Conference will focus on the development of scalable #ML platforms, #AIGovernance, and the development of #LLMs, #FederatedML, and Vision Transformers. The conference will feature expert-led talks, panels, and workshops designed to address both the technical nuances and the broader implications of #AI regulation for enterprises. Applications are now open for speakers: #MachineLearning #GenAI

📚 Another conference, another article! The International Conference on Machine Learning (ICML) took place last week, from July 21 to 27, 2024, in Vienna. In this article I explore some exciting advances in Machine Learning presented at this year’s edition: 🔍 For knowledge transfer, I highlight LLM distillation, 2-bit quantization, and transfer learning, which aim to create smaller and efficient models. 📈 In time series analysis, I explore zero-shot forecasting and metadata-enhanced time series generation, which is especially useful in energy field. Discover more in my medium article!

Machine learning algorithms are not only transforming entire industries but also reshaping our approach to problem-solving itself. In today’s CogX newsletter, Anil Ananthaswamy, an award-winning science writer and the author of “Why Machines Learn: The Elegant Maths Behind Modern AI,” questions the oversimplified view that reduces machine learning to mere "glorified statistics". Explore more at: https://lnkd.in/erHVyA_n

A very recent preprint to share—fantastic work by Kensuke Arai, Suma Jacob, MD, PhD, and Alik Widge! In this research, we introduce an experimental framework leveraging a game of iterated Rock-Scissors-Paper against an AI agent. The study employs a behavioral hypothesis of rule-switching to motivate analytical methods that extract behavioral features from game data. https://lnkd.in/e4C-AK3k This aligns closely with our other research findings: instead of focusing on average accuracy or reaction time in behavioral data, we investigate patterns of behavior to probe encoding mechanisms and/or personal traits.

Kudos Han Z. (another phenomenal PhD student from my team!) for running at the forefront of innovation! Don't miss my keynote at IEEE ICC'24 for a sneak peak into the future where I will speak about our latest GenAI research -> https://lnkd.in/geHCU532 In our paper "Generative AI-in-the-loop: Integrating LLMs and GPTs into the Next Generation Networks" (https://lnkd.in/gk3ANqrt) we are looking into integrating the super intelligence of GenAI into augmenting humans as well as machine learning algorithms that are used in networks. "Generative AI-in-the-loop" uses the semantic understanding, context awareness, and reasoning abilities of LLMs to assist orchestrating and decision making. We also thank our collaborators Akram Bin Sediq, PhD and Ali Afana for their contributions.

Convergence 2024 is on the horizon and we're looking for speakers! This May 8-9, the Convergence Conference will focus on the development of scalable #ML platforms, #AIGovernance, and the development of #LLMs, #FederatedML, and Vision Transformers. The conference will feature expert-led talks, panels, and workshops designed to address both the technical nuances and the broader implications of #AI regulation for enterprises. Applications are now open for speakers: #MachineLearning #GenAI