2 citations · 4 across the 7 of their papers we have counts for
7 papers
Emergent World Beliefs: Exploring Transformers in Stochastic Games
Adam Kamel, Tanish Rastogi, Michael Ma +2
Transformer-based large language models (LLMs) have demonstrated strong reasoning abilities across diverse fields, from solving programming challenges to competing in strategy-inte…
Automated Reward Design for Gran Turismo
Michel Ma, Takuma Seno, Kaushik Subramanian +3
When designing reinforcement learning (RL) agents, a designer communicates the desired agent behavior through the definition of reward functions - numerical feedback given to the a…
Do Transformer World Models Give Better Policy Gradients?
Michel Ma, Tianwei Ni, Clement Gehring +2
A natural approach for reinforcement learning is to predict future rewards by unrolling a neural network world model, and to backpropagate through the resulting computational graph…
Bridging State and History Representations: Understanding Self-Predictive RL
Tianwei Ni, Benjamin Eysenbach, Erfan Seyedsalehi +4
Representations are at the core of all deep reinforcement learning (RL) methods for both Markov decision processes (MDPs) and partially observable Markov decision processes (POMDPs…
Semi-Supervised Learning with Multiple Imputations on Non-Random Missing Labels
Jason Lu, Michael Ma, Huaze Xu +1
Semi-Supervised Learning (SSL) is implemented when algorithms are trained on both labeled and unlabeled data. This is a very common application of ML as it is unrealistic to obtain…
When Do Transformers Shine in RL? Decoupling Memory from Credit Assignment
Tianwei Ni, Michel Ma, Benjamin Eysenbach +1
Reinforcement learning (RL) algorithms face two distinct challenges: learning effective representations of past and present observations, and determining how actions influence futu…
SONAR: Joint Architecture and System Optimization Search
Elias Jääsaari, Michelle Ma, Ameet Talwalkar +1
There is a growing need to deploy machine learning for different tasks on a wide array of new hardware platforms. Such deployment scenarios require tackling multiple challenges, in…