Unverified50% confidenceFactExact time
Integrated gradients is an attribution method that assigns importance by integrating gradients along the path from a baseline to the input.
1
Sources
50%
Confidence
Long-term
Relevance
10/13/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn automatic differentiation, when a variable affects the loss function through multiple computation paths, the gradients from each path must be summed (gradient accumulation).73% similarUnverified加法节点的梯度传播规则是L对任一加数的偏导为1，直接将全局梯度传下去；乘法节点的偏导等于另一因子的值再乘以全局梯度71% similarUnverifiedThe Pulling step moves the embedding representation toward the weighted average of its neighbors, and the Correcting step projects the result back onto the learned manifold via the autoencoder.68% similarUnverifiedIn the context of reinforcement learning, a rollout refers to the process of an agent completing one full pass through an environment or policy to generate a single trajectory.67% similarUnverifiedBilinear interpolation is differentiable with respect to offsets, allowing gradients of sampling positions to be propagated through backpropagation, which is the mathematical foundation for end-to-end training of deformable attention.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1020479API
curl https://kongchang.com/api/v1/knowledge/claims/1020479MCP
get_claim(id=1020479)