greece
Algorithms
Data Science
Artificial Intelligence
Presenters: Yao Xie, Taiji Suzuki, and Xiuyuan Cheng
As a paradigm for sequential decision making in unknown environments,
reinforcement learning (RL) has received a flurry of attention in recent years. However, the explosion of model complexity in emerging applications and the presence of nonconvexity exacerbate the challenge of achieving efficient RL in sample-starved situations,
where data collection is expensive, time-consuming, or even high-stakes (e.g., in clinical trials, autonomous systems, and online advertising). How to understand and enhance the sample and computational efficiencies of RL algorithms is thus of great interest
and in imminent need. In this tutorial, we aim to present a coherent framework that covers important algorithmic and information-theoretic developments in RL, highlighting the connections between new ideas and classical topics. Employing Markov Decision Processes
as the central mathematical model, we start by introducing classical dynamic programming algorithms when precise descriptions of the environments are available. Equipped with this preliminary background, we introduce four distinctive RL scenarios (i.e., RL
with a generative model, online RL, offline RL, and multi-agent RL), and present three mainstream RL paradigms (i.e., model-based approach, model-free approach, and policy optimization). Our discussions gravitate around two central questions: what are the
information-theoretic limits in each RL scenario, and can we achieve these fundamental limits using computationally efficient algorithms? We will systematically introduce several effective algorithmic ideas (e.g., the optimism principle for online RL, the
pessimism principle for offline RL, variance reduction) that permeate the design of efficient RL algorithms.
Presenters: Ahmad Beirami and Ananda Theertha Suresh
In recent years, large language models have been used to solve a multitude of natural language tasks. In this tutorial, we give a brief overview of the history of language modeling and the fundamental techniques that led to the development of the modern language models behind Claude, Gemini, GPT, and Llama. We then present an overview of inference techniques for language models, when they are treated as black-box probabilistic models.
The tutorial is structured as four segments: In the first segment, we provide a historical perspective on language models and present a black-box probabilistic model that serves as a foundation for understanding the inference from these models. In the second segment, we examine black-box techniques aimed at aligning models towards various goals (e.g., safety), such as controlled decoding and the best-of-N algorithm. In the third segment, our focus then shifts to efficiency, where we examine information-theoretic techniques designed to improve inference latency, such as model compression or speculative decoding. If time permits, in the final segment, we discuss in-context learning, which has become one of the standard ways to solve new tasks using black-box language models.
Throughout the tutorial, we highlight open problems that might be of interest to the audience. The tutorial does not assume any prior knowledge on language models.
Presenters: Hye Won Chung and Lele Wang
Presenters: Rashmi Vinayak and Saurabh Kadekodi
By 2025, the total data stored in the cloud is estimated to grow over 100 Zettabytes (1 ZB = 1 million TB). Given that Moore’s Law and Kryder’s Law (the storage equivalent of Moore’s Law) have both ended, thus ending the “free” scaling benefits of processing power and storage capacity, the only way to survive the data juggernaut is by building larger-and-larger distributed systems. Modern exascale storage clusters are made up of hundreds-of-thousands of storage devices. At such scales failures are common, and erasure coding is the de-facto redundancy mechanism for protection against permanent data loss. Despite the widespread adoption of erasure codes in storage clusters and the advances in coding theory, the scale and resource-efficiency challenges posed by modern exascale clusters necessitate radically new ideas and innovation.
There is a complex balancing act between various resources which drive the cost of running large-scale storage clusters. The usual culprits are storage capacity, spindle (disk-time) efficiency, network traffic, memory utilization and computation. This tutorial will introduce the complex systems tradeoffs in an accessible way to the information-theory community, viewing the coding challenges in modern storage clusters from a practical lens. First, we describe the challenges, goals and techniques of erasure coding done in current exascale storage systems. Second, we provide a data-driven motivation for why one-scheme-fits-all erasure coding approaches employed today are both inefficient and insufficient. Third, we describe the systems architecture for designing adaptable storage systems, and cover the state-of-the-art theoretical models and code constructions that enable adaptability. Fourth, we showcase the technique of adding a new erasure code in a representative, popularly used open-source distributed storage system. Finally we end the tutorial by discussing open research directions. Overall, this tutorial will provide a holistic understanding of the setup, challenges, existing solutions and open research directions in coding practices of modern exascale storage systems.
Presenters: Volkan Cevher, Grigorios Chrysos, Fanghui Liu, and Leena Chennuru Vankadara
Historically, we are thought to use concise signal models in information theory, signal
processing, and machine learning to generalize better in the wild. Counter intuitively, this dogma has been proven (partly) wrong by deep neural networks, while rising into stardom, e.g., large video/language models, widely used for natural language, image,
speech, and other multi-modal tasks with unprecedented performance. In this tutorial, we will cover foundational deep learning theory that explains their scaling behavior, identify fundamental flaws in setting up learning goals that govern their robustness,
and the impact of the architectural choices into robustness.
To position Greece as a leading player in AI and Data Science
To build an AI Excellence Hub in Greece where the international research community can connect, groundbreaking ideas can thrive, and the next generation of scientists emerges, shaping a brighter future for Greece and the world
Welcome to ARCHIMEDES, a vibrant research hub connecting the global AI and Data Science research community fostering groundbreaking research in Greece and beyond. Its dedicated core team, comprising lead researchers, affiliated researchers, Post-Docs, PhDs and interns, is committed to advancing basic and applied research in Artificial Intelligence and its supporting disciplines, including Algorithms, Statistics, Learning Theory, and Game Theory organized around 6 core research areas. By collaborating with Greek and Foreign Universities and Research Institutes, ARCHIMEDES disseminates its research findings fostering knowledge exchange and providing enriching opportunities for students. Leveraging AI to address real-world challenges, ARCHIMEDES promotes innovation within the Greek ecosystem and extends its societal impact. Established in January 2022, as a research unit of the Athena Research Center with support from the Committee Greece 2021, ARCHIMEDES is funded for its first four years by the EU Recovery and Resilience Facility (RRF).
ARCHIMEDES, the AI and Data Science Research Hub at ATHENA Research Center, was thrilled to be part of the 2024 Athens Science Festival (ASF). With over 15,000 visitors and 8,000 students, the festival provided the opportunity for young and old, schools and families, and hundreds of visitors to explore science through fun, innovative, and interactive ways.
Στην υπογραφή Μνημονίου Συνεργασίας προχώρησαν στις 30 Ιουλίου 2024, η Πρόεδρος της Επιτροπής Κεφαλαιαγοράς, κα Βασιλική Λαζαράκου και ο Πρόεδρος του Διοικητικού Συμβουλίου και Γενικός Διευθυντής του «ΑΘΗΝΑ» – Ερευνητικό Κέντρο Καινοτομίας στις Τεχνολογίες της Πληροφορίας, των Επικοινωνιών και της Γνώσης», κ. Ιωάννης Εμίρης.
The Archimedes Research Unit on Artificial Intelligence, Data Science, and Algorithms proudly organized the Workshop on the Foundations of Modern AI at the National Technical University of Athens, Greece, on July 3-4, 2024. This prestigious event gathered leading experts and researchers from diverse fields such as Statistical and Computational Learning Theory, Optimization, Algorithmic Game Theory, and Approximation and Online Algorithms.
The appointed new Deputy Minister for Development, Ms. Zoe Rapti, visited on Thursday, July 4, 2024, the Athena Research Center in Athens. The President of Athena, Professor Ioannis Emiris, and Directors and Representatives of the Institutes and Units of Athena received Deputy Minister Zoe Rapti at the entrance of the General Secretariat of Research and Technology. Some of the facilities available at Athena were shown to her, and she briefly interacted with some researchers and staff.
The British Heart Foundation (BHF) has awarded £4 million to support world-class cardiovascular research at The University of Manchester over the next five years, with the university matching this funding to bring the total investment to £8 million. Archimedes is proud to announce that one of our lead researchers, Professor Sophia Ananiadou of Manchester University, is co-leading this transformative project.
The project “ARCHIMEDES Unit: Research in Artificial Intelligence, Data Science and Algorithms” with code OPS 5154714 is implemented by the National Recovery and Resilience Plan “Greece 2.0” and is funded by the European Union – NextGenerationEU.
Stay connected! Subscribe to our mailing list by emailing sympa@lists.athenarc.gr
with the subject "subscribe archimedes-news Firstname LastName"
(replace with your details)
