Expired50% confidenceBenchmarkExact time
PPO's held-out set economic return plateaus between approximately 50,000 and 300,000 steps, and is notably lower than a well-tuned rule-based baseline.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/19/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/17/2026(expired)
Sources
温室控制器大比拼：强化学习为何败给规则基线？
Related Claims
Cite This Claim
Stable URI
https://kongchang.com/claim/561525API
curl https://kongchang.com/api/v1/knowledge/claims/561525MCP
get_claim(id=561525)