Taylor Representations for Model-Free RL in Networked MDPs
arXiv:2610.05456v1 Announce Type: new Abstract: In Networked Markov Decision Processes, transition dynamics are often unknown and the state--action space grows rapidly with the number of agents. In this setting, Taylor representations naturally approximate $Q$-functions, but a naive order-$n$ expan…
Read the full story at arXiv stat.ML ↗
Timeline · 1 report
- 2026-10-06 04:00 · arXiv stat.ML
Taylor Representations for Model-Free RL in Networked MDPs