Authors
- Serena Booth*
- W. Bradley Knox*
- Julie Shah*
- Scott Niekum*
- Peter Stone
- Alessandro Allievi*
* External authors
Venue
- AAAI 2023
Date
- 2023
The Perils of Trial-and-Error Reward Design: Misdesign through Overfitting and Invalid Task Specifications
Serena Booth*
W. Bradley Knox*
Julie Shah*
Scott Niekum*
Alessandro Allievi*
* External authors
AAAI 2023
2023
Abstract
In reinforcement learning (RL), a reward function that aligns exactly with a task's true performance metric is often sparse. For example, a true task metric might encode a reward of 1 upon success and 0 otherwise. These sparse task metrics can be hard to learn from, so in practice they are often replaced with alternative dense reward functions. These dense reward functions are typically designed by experts through an ad hoc process of trial and error. In this process, experts manually search for a reward function that improves performance with respect to the task metric while also enabling an RL algorithm to learn faster. One question this process raises is whether the same reward function is optimal for all algorithms, or, put differently, whether the reward function can be overfit to a particular algorithm. In this paper, we study the consequences of this wide yet unexamined practice of trial-and-error reward design. We first conduct computational experiments that confirm that reward functions can be overfit to learning algorithms and their hyperparameters. To broadly examine ad hoc reward design, we also conduct a controlled observation study which emulates expert practitioners' typical reward design experiences. Here, we similarly find evidence of reward function overfitting. We also find that experts' typical approach to reward design---of adopting a myopic strategy and weighing the relative goodness of each state-action pair---leads to misdesign through invalid task specifications, since RL algorithms use cumulative reward rather than rewards for individual state-action pairs as an optimization target. Code, data: https://github.com/serenabooth/reward-design-perils.
Related Publications
Having explored an environment, intelligent agents should be able to transfer their knowledge to most downstream tasks within that environment. Referred to as ``zero-shot learning," this ability remains elusive for general-purpose reinforcement learning algorithms. While rec…
Scaling up the model size and computation has brought consistent performance improvements in supervised learning. However, this lesson often fails to apply to reinforcement learning (RL) because training the model on non-stationary data easily leads to overfitting and unstab…
Deep reinforcement learning has achieved superhuman racing performance in high-fidelity simulators like Gran Turismo 7 (GT7). It typically utilizes global features that require instrumentation external to a car, such as precise localization of agents and opponents, limiting …
JOIN US
Shape the Future of AI with Sony AI
We want to hear from those of you who have a strong desire
to shape the future of AI.