Curious Replay for DreamerV3
What it is
A JAX implementation of the Curious Replay method integrated into the DreamerV3 model-based reinforcement learning agent. It prioritizes experience replay by sampling past experiences that are most unfamiliar or surprising to the agent, using curiosity signals to keep the world model up-to-date in changing environments.
The problem it solves
Standard experience replay in model-based RL agents may not prioritize the most informative experiences, leading to slow or ineffective adaptation when the environment changes.
What you could use it for
- Training model-based RL agents in non-stationary environments
- Improving adaptation speed for DreamerV3 agents
- Benchmarking curiosity-driven replay strategies in Crafter and DMC tasks
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.