Unverified50% confidenceFactExact time
PPO的目标函数为L(θ) = E[min(r(θ)·A, clip(r(θ), 1-ε, 1+ε)·A)]，其中ε通常取0.1或0.2
1
Sources
50%
Confidence
Long-term
Relevance
8/5/2026
First Seen
Sources
Related Claims
Unverified贝叶斯定理表达为P(A|B) = P(B|A)·P(A)/P(B)61% similarUnverified导数的数值近似公式为 (f(a+h)-f(a))/h，源自微积分中极限的定义57% similarUnverifiedJohnson-Lindenstrauss引理保证对于任意ε∈(0,1)，只要目标维度k≥O(log(n)/ε²)，就存在保持两两距离在(1±ε)倍之内的高维到k维映射55% similarUnverified墨卡托投影的坐标变换公式为 x = R·λ，y = R·ln[tan(π/4 + φ/2)]54% similarUnverified对于满足拉普拉斯方程的调和函数，域内任意一点的值等于以该点为中心的球面上函数值的平均，这一性质被称为均值性质53% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/690916API
curl https://kongchang.com/api/v1/knowledge/claims/690916MCP
get_claim(id=690916)