PBV Method for Offline RL Model Selection Draws Peer Scrutiny

Van der Laan's team proposed Pairwise Bellman Validation (PBV), a tuning-free method with oracle guarantees for model selection in offline RL and OPE, prompting Nan Jiang's questions about its relation to LSTD tournaments and theoretical conditions.

2026-10-28 ~ 2026-10-29 · 3 related posts

• Pairwise Bellman Validation: tuning-free model selection for offline RL and OPE — LarsvanderLaan3 · 2026-10-28
• Researcher probes new tuning-free offline RL model selection method's LSTD ties — nanjiang_cs · 2026-10-29
• Follow-up: does offline RL estimator selection need singular value bounds? — nanjiang_cs · 2026-10-29