Unverified50% confidenceFactExact time
A 2018 study by the Google Brain team showed that simply changing the random seed could cause performance fluctuations of several percentage points in certain models.
1
Sources
50%
Confidence
Long-term
Relevance
8/17/2026
First Seen
Sources
Hugging Face复现2200篇ICML论文：AI研究可复现性危机全面曝光
Related Claims
Unverified在某些任务上，仅因随机种子不同导致的性能波动幅度可能超过某些论文声称的方法创新带来的提升76% similarUnverified即便固定所有随机种子，在不同硬件或并行度下，训练结果仍可能出现被大模型放大为显著能力差异的细微偏差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72% similarUnverifiedIn few-shot prompting, simply changing the order of examples can cause dramatic fluctuations in accuracy, ranging from near-random to near-optimal.71% similarUnverified深度学习中的随机种子控制权重初始化、数据加载批次打乱顺序、Dropout神经元随机失活以及优化器中的随机梯度估计等多个随机性来源68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/762110API
curl https://kongchang.com/api/v1/knowledge/claims/762110MCP
get_claim(id=762110)