Unverified50% confidenceFactExact time
In the stochastic-curvature quadratic model, there is a substantial trade-off between the rate of forgetting the initialization and the steady-state variance: faster forgetting is usually accompanied by larger steady-state variance.
1
Sources
50%
Confidence
Long-term
Relevance
10/14/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedNumerical discretization error arises from decomposing the theoretically continuous reverse-time stochastic differential equation process into a finite number of discrete time steps for execution, with fewer steps leading to larger per-step approximation errors.77% similarUnverifiedThe two core metrics for measuring machine unlearning quality are Forget Rate and Retain Rate; these two metrics are in natural tension, as excessive forgetting can lead to concept collapse.74% similarVerifiedThe Ebbinghaus forgetting curve reveals that memory strength decays exponentially over time, and this process can be slowed through spaced repetition.74% similarUnverified波动率具有长记忆特性，其自相关函数衰减极为缓慢，Hurst指数通常在0.7-0.9之间，远高于随机游走的0.574% similarUnverified权重初始化的随机性对模型性能影响最深远，数据批次打乱顺序的随机性影响最小73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1023469API
curl https://kongchang.com/api/v1/knowledge/claims/1023469MCP
get_claim(id=1023469)