World Wires · story 45390 · corroborated · 1 source(s)

Average-Reward Policy Identification from Single Trajectories

This paper analyzes the sample complexity and efficiency of identifying average-reward policies from a single continuing trajectory.

Open in the desk

Coverage

• Average-Reward Policy Identification from Single Trajectories arXiv: new preprints (security, AI, physics) · tier C · clear net · 2026-10-14 16:57 UTC

What this site indexes