Mastering diverse domains through world models

D Hafner, J Pasukonis, J Ba, T Lillicrap - ar** a general algorithm that learns to solve tasks across a wide range of
applications has been a fundamental challenge in artificial intelligence. Although current …

Learning agile soccer skills for a bipedal robot with deep reinforcement learning

T Haarnoja, B Moran, G Lever, SH Huang… - Science Robotics, 2024 - science.org
We investigated whether deep reinforcement learning (deep RL) is able to synthesize
sophisticated and safe movement skills for a low-cost, miniature humanoid robot that can be …

A practical guide to multi-objective reinforcement learning and planning

CF Hayes, R Rădulescu, E Bargiacchi… - Autonomous Agents and …, 2022 - Springer
Real-world sequential decision-making tasks are generally complex, requiring trade-offs
between multiple, often conflicting, objectives. Despite this, the majority of research in …

Rewarded soups: towards pareto-optimal alignment by interpolating weights fine-tuned on diverse rewards

A Rame, G Couairon, C Dancette… - Advances in …, 2024 - proceedings.neurips.cc
Foundation models are first pre-trained on vast unsupervised datasets and then fine-tuned
on labeled data. Reinforcement learning, notably from human feedback (RLHF), can further …

[HTML][HTML] dm_control: Software and tasks for continuous control

S Tunyasuvunakool, A Muldal, Y Doron, S Liu, S Bohez… - Software Impacts, 2020 - Elsevier
The dm_control software package is a collection of Python libraries and task suites for
reinforcement learning agents in an articulated-body simulation. Infrastructure includes a …

Challenges of real-world reinforcement learning: definitions, benchmarks and analysis

G Dulac-Arnold, N Levine, DJ Mankowitz, J Li… - Machine Learning, 2021 - Springer
Reinforcement learning (RL) has proven its worth in a series of artificial domains, and is
beginning to show some successes in real-world scenarios. However, much of the research …

Multi-objective gflownets

M Jain, SC Raparthy… - International …, 2023 - proceedings.mlr.press
We study the problem of generating diverse candidates in the context of Multi-Objective
Optimization. In many applications of machine learning such as drug discovery and material …

Pareto set learning for expensive multi-objective optimization

X Lin, Z Yang, X Zhang… - Advances in neural …, 2022 - proceedings.neurips.cc
Expensive multi-objective optimization problems can be found in many real-world
applications, where their objective function evaluations involve expensive computations or …

Scalar reward is not enough: A response to silver, singh, precup and sutton (2021)

P Vamplew, BJ Smith, J Källström, G Ramos… - Autonomous Agents and …, 2022 - Springer
The recent paper “Reward is Enough” by Silver, Singh, Precup and Sutton posits that the
concept of reward maximisation is sufficient to underpin all intelligence, both natural and …

An empirical investigation of the challenges of real-world reinforcement learning

G Dulac-Arnold, N Levine, DJ Mankowitz, J Li… - arxiv preprint arxiv …, 2020 - arxiv.org
Reinforcement learning (RL) has proven its worth in a series of artificial domains, and is
beginning to show some successes in real-world scenarios. However, much of the research …