Capability

Adaptive Offline Online Sample Mixing With Importance Weighting

1 artifact provides this capability.

Want a personalized recommendation?

Top Matches

Efficient Online Reinforcement Learning with Offline Data (RLPD)Product21/100

via “adaptive offline-online sample mixing with importance weighting”

* ⏫ 03/2023: [Reward Design with Language Models](https://arxiv.org/abs/2303.00001)

Unique: RLPD's adaptive weighting mechanism uses divergence-based feedback to automatically adjust offline-online ratios, whereas prior work (AWR, CQL) uses fixed ratios or manual scheduling. This enables the algorithm to gracefully transition from offline-dominated to online-dominated learning as the policy improves.

Adaptive Offline Online Sample Mixing With Importance Weighting

Top Matches

Also Known As

Company