NeurIPS 2023 Poster Session 1 - Adversarial Patches, Continual Learning, and Reinforcement Learning

Explore a comprehensive overview of four cutting-edge research papers presented at the NeurIPS 2023 Poster Session 1. Dive into topics including evolutionary strategies for generating camouflaged adversarial patches, exploiting heterogeneity in class distributions for exemplar-free continual learning, causes and effects of unanticipated numerical deviations in neural network inference frameworks, and conservative state value estimation for offline reinforcement learning. Gain insights into these advanced machine learning concepts through this informative 19-minute video presentation by Yannic Kilcher, offering a concise yet thorough examination of each paper's key findings and contributions to the field.