Explore, Exploit or Listen: Combining Human Feedback and Policy Model to Speed up Deep Reinforcement Learning in 3D Worlds
On the map: LLM alignment / RLHF / preference optimisation → Reward modeling and RLHF
Abstract (arXiv)
We describe a method to use discrete human feedback to enhance the performance of deep learning agents in virtual three-dimensional environments by extending deep-reinforcement learning to model the confidence and consistency of human feedback. This enables deep reinforcement learning algorithms to determine the most appropriate time to listen to the human feedback, exploit the current policy model, or explore the agent's environment. Managing the trade-off between these three strategies allows DRL agents to be robust to inconsistent or intermittent human feedback. Through experimentation using a synthetic oracle, we show that our technique improves the training speed and overall performance of deep reinforcement learning in navigating three-dimensional environments using Minecraft. We further show that our technique is robust to highly innacurate human feedback and can also operate when no human feedback is given.
Key facts. Cited by 8 papers in the map (data through 14 Oct 2026). Placed in Reward modeling and RLHF, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 4 of its 20 references that are in the map. First posted on arXiv 12 Sep 2017.
Builds on
4 of 20 references.
Closest ancestors · 1
Same direction, newest first
Other references · 3
- 2013Playing Atari with Deep Reinforcement Learningcited by 3,274
- 2016Continuous Deep Q-Learning with Model-based Accelerationcited by 157
- 2016Control of Memory, Active Perception, and Action in Minecraftcited by 61
Built on by
Citing papers per year: 2018: 1; 2019: 0; 2020: 0; 2021: 1; 2022: 2; 2023: 2; 2024: 2.
8 papers.
- 2018Reward learning from human preferences and demonstrations in Ataricited by 316
- 2021Training Value-Aligned Reinforcement Learning Agents Using a Normative Priorcited by 14
- 2024Mapping out the Space of Human Feedback for Reinforcement Learning: A Conceptual Frameworkcited by 5
- 2023A Framework for Learning from Demonstration with Minimal Human Effortcited by 4
- 2022A Brief Guide to Designing and Evaluating Human-Centered Interactive Machine Learningcited by 2
- 2022Reinforcement Learning for UAV control with Policy and Reward Shapingcited by 1
- 2023Reinforcement Learning With Reward Machines in Stochastic Gamescited by 0
- 2024Beyond Following: Mixing Active Initiative into Computational Creativitycited by 0
- 2024Mapping out the Space of Human Feedback for Reinforcement Learning: A Conceptual Framework18 Nov 2024
- 2024Beyond Following: Mixing Active Initiative into Computational Creativity6 Sep 2024
- 2023A Framework for Learning from Demonstration with Minimal Human Effort15 Jun 2023
- 2023Reinforcement Learning With Reward Machines in Stochastic Games27 May 2023
- 2022Reinforcement Learning for UAV control with Policy and Reward Shaping6 Dec 2022
- 2022A Brief Guide to Designing and Evaluating Human-Centered Interactive Machine Learning20 Apr 2022
- 2021Training Value-Aligned Reinforcement Learning Agents Using a Normative Prior19 Apr 2021
- 2018Reward learning from human preferences and demonstrations in Atari15 Nov 2018
Similar, not cited
Close in wording, no citation either way
- 2019Deep Reinforcement Learning from Policy-Dependent Human Feedback
- 2024A Survey On Enhancing Reinforcement Learning in Complex Environments: Insights from Human and LLM Feedback
- 2022Safe Deep RL in 3D Environments using Human Feedback
- 2025An Introduction to Deep Reinforcement and Imitation Learning
- 2024GUIDE: Real-Time Human-Shaped Agents
- 2026Themis: An explainable AI-enabled framework for Reinforcement Learning with Human Feedback
- 2018DQN-TAMER: Human-in-the-Loop Reinforcement Learning with Intractable Feedback
- 2023Breadcrumbs to the Goal: Goal-Conditioned Exploration from Human-in-the-Loop Feedback