254 citations · 312 across the 10 of their papers we have counts for
12 papers
Tokenization counts: the impact of tokenization on arithmetic in frontier LLMs
Aaditya K. Singh, DJ Strouse
Tokenization, the division of input text into input tokens, is an often overlooked aspect of the large language model (LLM) pipeline and could be the source of useful or harmful in…
Confronting Reward Model Overoptimization with Constrained RLHF
Ted Moskovitz, Aaditya K. Singh, DJ Strouse +4
Large language models are typically aligned with human preferences by optimizing (RMs) fitted to human feedback. However, human preferences are multi-facet…
Melting Pot 2.0
John P. Agapiou, Alexander Sasha Vezhnevets, Edgar A. Duéñez-Guzmán +14
Multi-agent artificial intelligence research promises a path to develop intelligent technologies that are more human-like and more human-compatible than those produced by "solipsis…
In-context Reinforcement Learning with Algorithm Distillation
Michael Laskin, Luyu Wang, Junhyuk Oh +11
We propose Algorithm Distillation (AD), a method for distilling reinforcement learning (RL) algorithms into neural networks by modeling their training histories with a causal seque…
Semantic Exploration from Language Abstractions and Pretrained Representations
Allison C. Tam, Neil C. Rabinowitz, Andrew K. Lampinen +6
Effective exploration is a challenge in reinforcement learning (RL). Novelty-based exploration methods can suffer in high-dimensional state spaces, such as continuous partially-obs…
Collaborating with Humans without Human Data
DJ Strouse, Kevin R. McKee, Matt Botvinick +2
Collaborating with humans requires rapidly adapting to their individual strengths, weaknesses, and preferences. Unfortunately, most standard multi-agent reinforcement learning tech…
Learning more skills through optimistic exploration
DJ Strouse, Kate Baumli, David Warde-Farley +2
Unsupervised skill learning objectives (Gregor et al., 2016, Eysenbach et al., 2018) allow agents to learn rich repertoires of behavior in the absence of extrinsic rewards. They wo…
Learning Truthful, Efficient, and Welfare Maximizing Auction Rules
Andrea Tacchetti, DJ Strouse, Marta Garnelo +2
From social networks to supply chains, more and more aspects of how humans, firms and organizations interact is mediated by artificial learning agents. As the influence of machine…
Social Influence as Intrinsic Motivation for Multi-Agent Deep Reinforcement Learning
Natasha Jaques, Angeliki Lazaridou, Edward Hughes +5
We propose a unified mechanism for achieving coordination and communication in Multi-Agent Reinforcement Learning (MARL), through rewarding agents for having causal influence over…
The information bottleneck and geometric clustering
DJ Strouse, David J Schwab
The information bottleneck (IB) approach to clustering takes a joint distribution and maps the data to cluster labels which retain maximal information…
The deterministic information bottleneck
DJ Strouse, David J Schwab
Lossy compression and clustering fundamentally involve a decision about what features are relevant and which are not. The information bottleneck method (IB) by Tishby, Pereira, and…
Levinson's theorem for graphs
Andrew M. Childs, DJ Strouse
We prove an analog of Levinson's theorem for scattering on a weighted (m+1)-vertex graph with a semi-infinite path attached to one of its vertices. In particular, we show that the…