Search papers, labs, and topics across Lattice.
Rochester Institute of Technology
2
0
3
Correcting and reweighting historical observations in non-stationary bandits can yield substantial sample efficiency gains, outperforming conventional methods that simply forget past data.
By federating distributional critics and using a Wasserstein barycenter trust region, TR-FedDistRL avoids the dangerous "mean-smearing" that can make federated RL unsafe in critical applications.
