Unverified50% confidenceSolutionExact time
作者在决策中加入少量随机化成分以避免行为过于可预测
1
Sources
50%
Confidence
Long-term
Relevance
7/20/2026
First Seen
Sources
Related Claims
UnverifiedIn few-shot prompting, simply changing the order of examples can cause dramatic fluctuations in accuracy, ranging from near-random to near-optimal.74% similarUnverified多次采样取多数票或自洽性方法可在推理层降低单次随机性对最终结果的影响74% 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67% similarUnverifiedExcessive domain randomization can cause policies to become overly conservative, and certain real-world phenomena (such as deformation of flexible objects) are difficult to cover through simple parameter randomization.65% similarUnverifiedEvery character in a prompt affects the probability distribution of all subsequent tokens, which is why subtle wording changes can lead to drastically different outputs.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/565527API
curl https://kongchang.com/api/v1/knowledge/claims/565527MCP
get_claim(id=565527)