MARL in real scenarios is still challenging due to the same safety and efficiency concerns in single-agent setting, then it is worth conducting investigation for offline RL in multi-agent setting.

Both algorithms yield state-of-the-art performance guarantees.

This bound guides the design of an optimal exploration policy attaining minimal sample complexity. However, this lower bound involves solving a hard non-convex optimization problem.

The latter case can be further reduced to adversarial MDP when preferences only depend on the final state.

The code is available in https://github.com/cathyhxh/DIFFER.

Unfortunately, choosing the best OPE algorithm for each task and domain is still unclear.