ML in PL

Regular Registration for ML in PL Conference ‘24 Opens on 9th September at 18:00 CEST (GMT+2)! Regular registration is on a first-come, first-served basis and will remain open until all tickets are sold out. Ticket prices: 👉 Student: 450 PLN 👉 Regular: 900 PLN Tickets can be purchased using this link: https://lnkd.in/d2zQWC9W

We’re incredibly grateful to our media partners for helping us make the ML in PL Conference 2024 a truly memorable event! 🚀

We’re filled with gratitude for the people and partnerships that made ML in PL Conference 2024 possible—especially our honorary patrons. Your support has been invaluable in bringing our mission to life: promoting machine learning in Poland and beyond, building a thriving community, bridging academia and industry, and fostering a collaborative environment that empowers emerging talent and promotes ethical innovation. ⭐ Centrum Nauki Kopernik  ⭐ Uniwersytet Warszawski  ⭐ Politechnika Warszawska  ⭐ Faculty of Mathematics, Informatics and Mechanics, UW ⭐ Faculty of Mathematics and Information Science, WUT ⭐ Faculty of Electronics and Information Technology, WUT

Welcome to the final day of the ML in PL Conference 2024! Today’s lineup is nothing short of extraordinary. A glimpse of what’s happening on the 3rd Day: Morning Invited Talks ▪️ Prof Alejandro Frangi FREng from the University of Manchester presented on the potential of AI-enabled In-silico Trials in Medical Innovation. ▪️ Tom Rainforth from the University of Oxford discussed Modern Bayesian Experimental Design. 10:30 - 12:00 |  Poster Session 2 ☕️ 12:00 - 13:25 | Contributed Talks 🎤 📍Main Hall: ▪️ Finding NeMo: Localizing Neurons Responsible For Memorization in Diffusion Models by Franziska Boenisch (CISPA Helmholtz Center for Information Security) ▪️ CDI: Copyrighted Data Identification in Diffusion Models by Jan Dubiński (Warsaw University of Technology / IDEAS NCBR) ▪️ Global Counterfactual Directions by Bartłomiej Sobieski (MI2.ai / University of Warsaw) 📍Hall A features: ▪️ Accelerating Goal-Conditioned RL Algorithms and Research by Michał Bortkiewicz (Warsaw University of Technology) ▪️ Fine-tuning Reinforcement Learning Models is Secretly a Forgetting Mitigation Problem by Bartłomiej Cupiał (University of Warsaw / IDEAS NCBR) ▪️ AdaGlimpse: Active Visual Exploration with Arbitrary Glimpse Position and Scale by Adam Pardyl (IDEAS NCBR / Jagiellonian University) 📍Hall B features: ▪️ Fixed Points of Nonnegative Neural Networks by Tomasz Piotrowski (Nicolaus Copernicus University in Toruń) ▪️ Augmentation-aware Self-supervised Learning with Conditioned Projector by Marcin Przewięźlikowski (Jagiellonian University / IDEAS NCBR) ▪️ Meta-analysis of Bayesian Analyses by Omar Rivasplata (University of Manchester) 13:30 - 15:00 | Lunch🍴 15:00 - 16:00 | Sponsor Talks 🎤 📍Hall A features: ▪️ LLM Serving by Long Cheng (Google) ▪️ Revolutionizing IT Operations with AI: The Comarch Experience by Tomasz Sapiński (Comarch) 📍Hall B features: ▪️ Neuro-Symbolic AI for Applications in Architecture and Design of Virtual Worlds by Przemyslaw Musialski (IDEAS NCBR) ▪️ Leveraging Feature Store for High-Sparsity Recommendations in LOT Polish Airlines by Daniel Śliwiński and Patryk Radon (LOT Polish Airlines) 16:10 - 17:10 | Discussion Panel and Invited Talks 📍In the Main Hall, "AI in Medicine Panel" will take place. 🎙️🔬 📍Hall A features: ▪️ Computer Vision in the Age of LLMs by Lucas Beyer (Google DeepMind) 📍Hall B features: ▪️ Improving Foundation Models (with Academic Compute) by Yuki Asano (University of Technology Nuremberg) 17:15 - 17:45 | Closing Ceremony 🎉 🎨 ML in PL: AI Art Festival – Open daily until 11th November from 10:00 - 20:00 at the Academy of Fine Arts (📍Wybrzeże Kościuszkowskie 37). 6-minute walk from the conference venue. The exhibition is free of charge! We invite you to join us for the AI Art Festival’s awards ceremony, followed by an exhibition tour led by the organizers. (Wybrzeże Kościuszkowskie 37 at 19:15)

Tomorrow, on Day 3 of ML in PL Conference 2024, Przemyslaw Musialski takes the stage to present "Neuro-Symbolic AI for Applications in Architecture and Design of Virtual Worlds". 💡Artificial intelligence is rapidly evolving, offering new tools that blend human reasoning with machine learning. One exciting development is neuro-symbolic AI, a hybrid approach combining neural networks with rule-based, symbolic reasoning. This method bridges the flexibility of deep learning with the interpretability of structured logic, making it ideal for creative and complex fields like architecture and design. In this talk, he will discuss how neuro-symbolic models are applied to tasks such as facade reconstruction, where they can transform images into detailed, adaptable models for use in design. This approach offers both precision and creativity, enabling architects and designers to move seamlessly between images and structured designs. Through examples, we’ll explore the potential of neuro-symbolic AI to enhance workflows and provide powerful, intuitive tools for architects, designers, and researchers alike.  🎤 Przemyslaw Musialski is the leader of the research group Computer Graphics at IDEAS NCBR and an associate professor of Computer Science at New Jersey Institute of Technology, USA. His research spans geometric modeling, geometry processing, computational fabrication, and machine learning, focusing on creating algorithmic solutions for digital content generation. Before joining IDEAS NCBR, he led the Computational Fabrication group at Vienna University of Technology’s Center for Geometry and Computational Design, and previously, he held academic and research positions at VRVis Vienna, TU Vienna, Arizona State University, and King Abdullah University of Science and Technology. He holds an MSc degree in Media Systems Science from Bauhaus University Weimar and a PhD in Computer Science from the Vienna University of Technology. He is a member of the ACM SIGGRAPH and the EUROGRAPHICS association. ⏰ Saturday, November 8th, 15:00 in Hall B

Day 2 of the ML in PL Conference 2024 has arrived, and it’s off to an exciting start at the Copernicus Science Centre! Roberto Calandra presented "Perceiving, Understanding, and Interacting through Touch", highlighting the evolving role of tactile interaction in robotics. Meanwhile, Stanisław Jastrzębski shared stories from the journey of building molecule.one, in his talk "Candid Stories and Hard Lessons from our Journey Building an AI for Science Startup". We have a busy day ahead! Here’s what’s coming up: 10:35 AM - 12:00 PM Sessions 📍Main Hall: Talks by the Witold Lipski Award Laureates on adaptiveness in deep learning models, applied computer science practices in ML for medical and space projects, and dynamic shortest paths data structures. 📍Hall A featuring: ▪️ Exceeding Historical Exposure in Session-Based Recommender Systems by Klaudia Balcer ▪️ Multi-Armed Bandit Algorithms for Dynamic Decision Making by Tudor C. ▪️ From Theory to Practice: A Practitioner's Journey with Knowledge Graphs by Patryk Wielopolski 📍Hall B with: ▪️Deep Learning for Effective Analysis of High Content Screening by Adriana Borowa ▪️ Aggregated Attributions for Explanatory Analysis of 3D Segmentation Models by Maciej Chrabąszcz ▪️Towards Medical Foundation Model - A Unified Dataset for Pretraining Medical Imaging Models by Barbara Klaudel 12:00 - 12:20 PM Coffee Break ☕ 12:20 - 1:20 PM 📍Main Hall hosts the Career Paths in ML Panel 🎙️ 📍Hall A featuring Allegro on AlleNoise, a large-scale text classification benchmark dataset, by Alicja Rączkowska  📍Hall B featuring Careers in Quant Finance by Charles Martinez PhD and Scarlett Bailey 1:20 - 2:30 PM Lunch🍴 2:30 - 4:00 PM 📍Main Hall featuring: ▪️ Graph Neural Networks and Diffusion Model for RNA 3D Structure Prediction by Marek Justyna ▪️Fake it till you make it: Planning Chemical Syntheses for Drug Discovery by Krzysztof Maziarz ▪️ Uncertainty-Aware Self-Supervised Learning on Multi-Dimensional Time Series for Animal Behavior by France Rose 📍Hall A featuring: ▪️ How LLMs are Revolutionizing the Cybersecurity Field by Natasha Al-Khatib, PhD ▪️ Efficient Fine-Tuning of LLMs: Exploring PEFT Methods and LoRA-XS Insights by Klaudia Bałazy ▪️ Open LLMs are Necessary for Private Adaptations and Outperform their Closed Alternatives by Adam Dziedzic 📍Hall B featuring: ▪️ Personalization of Large-Scale Diffusion Models by Kamil Deja ▪️ Current Trends in Intrinsically Interpretable Deep Learning by Dawid Rymarczyk, PhD ▪️ Neural Rendering: The Future of 3D Modeling by Przemysław Spurek 4:10 - 5:10 PM 📍Main Hall hosts AI Safety Panel by ElevenLabs 🎙️ 📍Hall A:  Inductive Biases for Robot Reinforcement Learning by Jan Peters 📍Hall B:  RLHF as conditioning on human preferences by Tomek Korbak 5:10 - 6:40 PM Poster Session 1, featuring posters 1-14, 16-31 🧑🎨The AI Art Exhibition is open.📍Wybrzeże Kościuszkowskie 37

Welcome to Day 1 of the ML in PL Conference 2024—one of the year’s biggest and boldest scientific events! 🎉 Saying we’re pumped is putting it mildly! The morning began with students presenting their remarkable work, setting a high standard for the sessions to follow. Quick glance at the agenda: 12:15 - 13:15 | Keynote Talks 🌟 Iryna Gurevych presents: Towards Real-World Fact-Checking with Large Language Models.📍Main Hall  🌟 Bernardino Romera Paredes shares insights on Evolving Programs with LLMs.📍Halls A & B 13:15 - 14:45 | Lunch🍴 14:45 - 15:45 | Afternoon Talks and Panels 🎙️ AI in Law panel explores AI’s impact on the legal sector.📍Main Hall  🌟 Wojciech Samek talks on Explainable AI for LLMs.📍Halls A & B 15:45 - 16:15 | Coffee Break ☕ Great opportunity to connect with fellow ML/AI enthusiasts! 16:15 - 17:45 | AI Audio Challenge Final 🌟 AI Audio Challenge Finale – ElevenLabs announces the winners!📍Main Hall 19:00 - 00:00 | Conference Party  🥂 Day 1 ends with the official ML in PL Conference party. Badges are required for entrance.📍Bolek Pub & Restaurant When you get a chance, head over to the ML in PL: AI Art Festival, just across the street from the Copernicus Science Centre. Participation is free of charge for everyone, whether you’re attending the conference or not.📍Wybrzeże Kościuszkowskie 37/39 (floors -1 and 2), open daily from 10 AM to 6 PM until November 11. 🎨

Meet Long Cheng from Google Cloud Vertex AI! 🎤 Long Cheng is an Engineering Leader at Google Cloud Vertex AI - a unified machine learning (ML) platform that aims to streamline the entire ML workflow, making it easier for data scientists and ML engineers to build, deploy, and scale AI models. Vertex AI offers a comprehensive suite of tools to manage every stage of the ML development lifecycle, from data preparation and model training to deployment, monitoring, and maintenance. As the site lead for Google Cloud Vertex AI in Poland, Long spearheads the development of MLOps infrastructure (Model training pipelines, Model Garden, Fine-tuning and etc.) and AI serving infrastructure for both online and batch predictions of Gemini and other pretrained AI models, leveraging GPU and TPU acceleration. Long brings a wealth of experience in product strategy and cross-functional leadership, effectively guiding engineering and product management teams to achieve evolving business goals. Prior to that, he gained over 12 years of experience in the U.S. as a lead engineer with top tech companies such as Google, Oracle and Microsoft. 💡In his presentation "Scaling and Optimizing LLM Serving", Long will walks us through techniques for deploying LLMs effectively to meet demanding performance standards. He’ll cover key fundamentals like prefill, decoding, and memory optimization, and explore crucial metrics such as latency, throughput, and cost. The session will also compare popular model-serving frameworks like vLLM, Hugging Face Text Generation Inference, and Nvidia TensorRT-LLM, along with introducing optimization techniques such as quantization, batching, and advanced memory management strategies. 📅 Day 3: Saturday, 9th November 🕓 15:00 - 15:30 📍 Hall A, Copernicus Science Center

We invite you to an in-depth look into the world of Quantitative Finance with G-Research. 📈 Dr. Charles Martinez and Scarlett Bailey will share insights on G-Research’s quantitative activities, what they seek in future quants, and their recruitment process. 💼 🎤 Charles Martinez PhD is the Academic Relations Manager at G-Research. Charles completed a PhD in Phonon interactions in Gallium Nitride nanostructures at the University of Nottingham. Charles' previous role was as Elsevier's Key Account Manager, managing sales and renewals for the UK Russell Group institutions, Government and Funding body accounts, including being one of the negotiators in the UK ScienceDirect Read and Publish agreement. Since leaving Elsevier Charles is dedicated to forming beneficial partnerships between G-Research and Europe's top institutions, and is living in Cambridge, UK. 🎤 Scarlett Bailey is currently an intern in the Talent Acquisition department at G-Research, completing her placement year while studying for a BSc in Management at the University of Bath. As part of the team, Scarlett supports recruitment for the Summer Research Programme and other internship opportunities, with a focus on engaging emerging talent in quantitative research and technology. She also contributes to G-Research's global attraction events, connecting students with career opportunities across the company. 📅 Date: Friday, November 8 (Day 2) ⏰ Time: 12:20 - 13:20 📍 Location: Hall B 👇 Find the links in the comments.

Meet our next speaker - Stanisław Jastrzębski! Stanislaw Jastrzebski serves as the CTO and Chief Scientist at Molecule.one, a biotech startup in the drug discovery space. He is passionate about improving the fundamental aspects of deep learning and applying it to automate scientific discovery. He completed his postdoctoral training at New York University in deep learning. His PhD thesis was based on work on foundations of deep learning done during research visits at MILA (with Yoshua Bengio) and the University of Edinburgh (with Amos Storkey). He received his PhD from Jagiellonian University, advised by Jacek Tabor. Beyond academia, he gained industrial experience at Google, Microsoft and Palantir. In his scientific work, he has published at leading machine learning venues (NeurIPS, ICLR, ICML, JMLR, Nature SR). He is also actively contributing to the machine learning community as an Area Chair (most recently NeurIPS ’23) and as an Action Editor for TMLR. At Molecule.one, he leads technical teams working on software for synthesis planning based on deep learning, public data sources, and experiments from a highly automated laboratory.