Deep Learning Extravaganza 2023

Celebrating the most cutting-edge machine learning research of the Amsterdam AI Ecosystem

June 8, 2023

16:00-19:30

Lab42, Amsterdam Science Park

Come and deep dive into the state-of-the-art machine learning research that will be showcased at the NeurIPS Conference in New Orleans!

Event programme

16:10-17:00

Keynote Presentation

Prof. Jakub Tomczak (Eindhoven University)
Venue: L1.02

17:00-19:30

Poster Session

with bites & drinks
Venue: ground floor

Prof. Dr. Jakub Tomczak

Associate Professor PI of Generative AI Group

About the keynote speaker

Jakub M. Tomczak is an associate professor at the Eindhoven University of Technology (TU/e). Before joining the TU/e, he was an assistant professor of Artificial Intelligence at Vrije Universiteit Amsterdam. Formerly, he was a deep learning researcher (Engineer, Staff) in Qualcomm AI Research in Amsterdam, a Marie Sklodowska-Curie individual fellow in Prof. Max Welling’s group at the University of Amsterdam, and an assistant professor and a postdoc at the Wroclaw University of Technology. His main research interests include deep generative modeling, deep learning, and Bayesian inference, with applications to image processing, life sciences & chemistry, and medicine. He is the author of the book entitled “Deep Generative Modeling”. He has broad experience in consulting companies in logistics, banking, computer vision, and life sciences. He has over 12 years of experience working in academia and over 3 years working in and for the industry.

Is there a place for Representation Learning in Generative AI?
Deep Learning, Representation Learning, and Generative AI. These are three buzzwords that appear not only in research papers but also newspapers and social media. However, how are these three concepts related to each other? Generative AI is built on top of the success of Deep Learning, but is there a place for Representation Learning? This talk aims at entertaining the audience with a brief introduction to Generative AI and Representation Learning, and at indicating a possible way of marrying these two fields.

Posters showcased at the event

Reducing Predictive Feature Suppression in Resource-Constrained Contrastive Image-Caption Retrieval

Published at the Transactions on Machine Learning Research (TMLR) 2023

Rigid body flows for sampling molecular crystal structures

Published at the International Conference on Machine Learning (ICML) 2023

An Offline Metric for the Debiasedness of Click Models

Published at the SIGIR 2023

Self-Guided Diffusion Models

Published at the CVPR 2023

Memory-efficient NLLB-200: Language-specific Expert Pruning of a Massively Multilingual Machine Translation Model

Published at the Association for Computational Linguistics (ACL) 2023

3D Human Pose Estimation via Intuitive Physics

Published at the CVPR 2023

Interpretable Word Sense Representations via Definition Generation: The Case of Semantic Change Analysis

Published at the ACL 2023

MultiTabQA: Generating Tabular Answers for Multi-Table Question Answering

Published at the ACL 2023

Tailed U-net Multi-scale Music Representation Learning

Published at the International Society for Music Information Retrieval (ISMIR) 2023

Unlocking Slot Attention by Changing Optimal Transport Costs

Published at the ICML 2023

BISCUIT: Causal Representation Learning from Binary Interactions

Published at the Conference on Uncertainty in Artificial Intelligence (UAI) 2023

Injecting the BM25 Score as Text Improves BERT-Based Re-rankers

Published at the European Conference on Information Retrieval (ECIR) 2023

Safe Deployment for Counterfactual Learning to Rank with Exposure-Based Risk Minimization

Published at the SIGIR 2023

Graph Switching Dynamical Systems

Published at the ICML 2023

Scene-centric vs. Object-centric Image-Text Cross-modal Retrieval: A Reproducibility Study

Published at the ECIR 2023

Multi-objective optimization via equivariant deep hypervolume approximation

Published at the ICLR 2023

SuperDisco: Super-Class Discovery Improves Visual Recognition for the Long-Tail

Published at the CVPR 2023

Speaking the Language of Your Listener: Audience-Aware Adaptation via Plug-and-Play Theory of Mind

Published at the Findings of ACL 2023

Modelling Long Range Dependencies in N-D: From Task-Specific to a General Purpose CNN

Published at the ICLR 2023

Causal Representation Learning for Instantaneous and Temporal Effects in Interactive Systems

Published at the International Conference on Learning Representation (ICLR) 2023

sigmoidF1: A Smooth F1 Score Surrogate Loss for Multilabel Classification

Published at the TMLR 2023

Test of Time: Instilling Video-Language Models with a Sense of Time

Published at the Computer Vision and Pattern Recognition (CVPR) 2023

Meta Learning to Bridge Vision and Language Models for Multimodal Few-Shot Learning

Published at the ICLR 2023

E(n) Equivariant Message Passing Simplicial Networks

Published at the ICML 2023

Balancing Fairness and Efficiency in Transport Network Design through Reinforcement Learning

Published at the International Conference on Autonomous Agents and Multiagent Systems (AAMAS) 2023

MetaModulation: Learning Variational Feature Hierarchies for Few-Shot Learning with Fewer Tasks

Published at the ICML 2023

A Unified Generative Retriever for Knowledge-Intensive Language Tasks via Prompt Learning

Published at the SIGIR 2023

Uncertainty-Aware Multiple-Instance Learning for Reliable Classification: Application to Optical Coherence Tomography

Under review

In the same spirit of the successful NeurIPS Fest held in November 2022, the ELLIS unit Amsterdam and the Deep Thinking Hour are organizing a festive event to recognize and celebrate the most cutting-edge machine learning research of the Amsterdam AI Ecosystem. We will have speakers, followed by a poster presentation (light refreshment provided).