In this presentation, Sourav Panda presents Fast and Data Efficient RL from Pixels Using Non-Parametric Value Approximation, by Long et al. In it, the authors introduce a way to learn action selection in RL domains directly from pixel-level feature data that is highly...
In this presentation, Jeff Schulman presents We’re Still People and Not Only Emails That We Are Sending: Shared Cognition in Distributed Design Collaboration…, by Nguyen and Mougenot. In it, the authors examine the presence and suitability of a variety of...
In this presentation, Mitchell Hoesing presents Why Most Published Research Findings Are False, by Ioannidis. In it, the author constructs a simple mathematical model based on parameters of statistical analysis and draws charts that compute probabilities that findings...
In this presentation, Mitchell Hoesing presents Attention is All You Need, by Vaswani et al. This paper introduces the transformer architecture, which is current state of the art on a number of benchmark tasks. Prior to the transformer, sequence data would usually...
In this presentation, Dr. Jonathan Dodge presents Generative Agents: Interactive Simulacra of Human Behavior, by Park et al. The authors create a sandbox environment resembling The Sims, with agents interacting with each other. They find generative agents could...
In this presentation, Dr. Jonathan Dodge presents Grandmaster level in StarCraft II using multi-agent reinforcement learning, by Vinyals et al. In this paper, the authors introduce AlphaStar, which plays a complex real-time strategy game as well as very-top-quality...
In this presentation, Rupak Das covers Language Models Are Few-Shot Learners, by Brown et al. The paper introduces GPT-3 and describes tests for the model’s performance using a few shot setting and a dataset. Both methods have advantages and disadvantages, but...
In this presentation, Dr. Jonathan Dodge covers Neurosymbolic Learning with Formally Verified Exploration, by Anderson et al. The paper introduces the concept of formally verified exploration in neurosymbolic reinforcement learning. The main goal is to ensure a safe...
Jeff Schulman recently led an enlightening session in the Sequential Domain subgroup at PLAINTEXT, focusing on the paper “Designing Ecosystems of Intelligence from First Principles” by Karl J. Friston and colleagues. This paper delves into foundational...
Rupak Das recently presented a compelling video review of a noteworthy NLP paper, Evaluation of Fake News Detection with Knowledge-Enhanced Language Models (Whitehouse et al., 2022), at PLAINTEXT’s Natural Language Processing subgroup. His insightful explanation...
