Matthew Aitchison
YOU?
Author Swipe
View article: Evaluating Frontier Models for Dangerous Capabilities
Evaluating Frontier Models for Dangerous Capabilities Open
To understand the risks posed by a new AI system, we must understand what it can and cannot do. Building on prior work, we introduce a programme of new "dangerous capability" evaluations and pilot them on Gemini 1.0 models. Our evaluations…
View article: Learning Universal Predictors
Learning Universal Predictors Open
Meta-learning has emerged as a powerful approach to train neural networks to learn new tasks quickly from limited data. Broad exposure to different tasks leads to versatile representations enabling general problem solving. But, what are th…
View article: Distributional Bellman Operators over Mean Embeddings
Distributional Bellman Operators over Mean Embeddings Open
We propose a novel algorithmic framework for distributional reinforcement learning, based on learning finite-dimensional mean embeddings of return distributions. We derive several new algorithms for dynamic programming and temporal-differe…
View article: Language Modeling Is Compression
Language Modeling Is Compression Open
It has long been established that predictive models can be transformed into lossless compressors and vice versa. Incidentally, in recent years, the machine learning community has focused on training increasingly large and powerful self-sup…
View article: Atari-5: Distilling the Arcade Learning Environment down to Five Games
Atari-5: Distilling the Arcade Learning Environment down to Five Games Open
The Arcade Learning Environment (ALE) has become an essential benchmark for assessing the performance of reinforcement learning algorithms. However, the computational cost of generating results on the entire 57-game dataset limits ALE's us…
View article: An Agile New Research Framework for Hybrid Human-AI Teaming: Trust, Transparency, and Transferability
An Agile New Research Framework for Hybrid Human-AI Teaming: Trust, Transparency, and Transferability Open
We propose a new research framework by which the nascent discipline of human-AI teaming can be explored within experimental environments in preparation for transferal to real-world contexts. We examine the existing literature and unanswere…
View article: Curriculum Generation and Sequencing for Deep Reinforcement Learning in StarCraft II
Curriculum Generation and Sequencing for Deep Reinforcement Learning in StarCraft II Open
Reinforcement learning has proven successful in games, but suffers from long training times when compared to other forms of machine learning. Curriculum learning, an optimisation technique that improves a model's ability to learn by presen…
View article: Learning to Deceive in Multi-agent Hidden Role Games
Learning to Deceive in Multi-agent Hidden Role Games Open
View article: Do Game Bots Dream of Electric Rewards?
Do Game Bots Dream of Electric Rewards? Open
The purpose of this paper is to draw together theories, ideas, and observations related to rewards, motivation, and play to develop and question our understanding and practice of designing reward-based systems and technology. Our explorati…
View article: Optimal Use of Experience in First Person Shooter Environments
Optimal Use of Experience in First Person Shooter Environments Open
Although reinforcement learning has made great strides recently, a continuing limitation is that it requires an extremely high number of interactions with the environment. In this paper, we explore the effectiveness of reusing experience f…
View article: Effects of fluoridation on oral health and clinical guidelines for fluoride therapy
Effects of fluoridation on oral health and clinical guidelines for fluoride therapy Open
Presented to the 11th Annual Symposium on Graduate Research and Scholarly Projects (GRASP) held at the Heskett Center, Wichita State University, April 24, 2015.