Unverified50% confidenceFactExact time
The parameter-shift rule is a gradient estimation method commonly used when training variational quantum circuits; for an expectation value of the form cos(θ), its derivative can be obtained exactly by evaluating the circuit at θ±π/2.
1
Sources
50%
Confidence
Long-term
Relevance
10/21/2026
First Seen
Sources
Related Entities
Related Claims
Unverified重参数化技巧将采样表示为 z = μ + σ × ε（ε从标准正态分布采样），使随机性外置于ε，从而让μ和σ的梯度可正常反向传播实现端到端训练62% similarUnverifiedRule drift refers to the phenomenon in which the underlying decision rules of an environment change non-explicitly over time; similar to concept drift, it requires the learning system to detect distribution changes and trigger model updates.62% similarUnverified大模型中的参数在数学上是神经网络每条连接边上的权重数值和偏置值，训练前随机初始化，通过反向传播算法和梯度下降不断调整62% similarUnverifiedGradient descent updates parameters along the opposite direction of the loss function's gradient, and the parameter update formula is usually w ← w − η·∂L/∂w, where η is the learning rate.61% similarUnverifiedThe judge's model version and temperature should be fixed, because a silently changing judge is a moving ruler; after each change to the judge prompt or judge model, the agreement rate should be re-checked.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1079384API
curl https://kongchang.com/api/v1/knowledge/claims/1079384MCP
get_claim(id=1079384)