相信我，我知道路：捷径学习存在下的预测不确定性
Trust Me, I Know the Way: Predictive Uncertainty in the Presence of Shortcut Learning
浏览论文内容
中文总结 AI 辅助
本文探讨了神经网络预测不确定性量化中关于熵分解与认知不确定性的争论，指出无知与分歧观点均成立但源于不同学习情境，并证明捷径的存在是认知不确定性表现为分歧的决定性因素。
中文摘要 AI 辅助
神经网络中预测不确定性的正确量化方法仍备受讨论。特别是，在将无知与分歧观点对立的争论下，目前最先进的熵分解是否能产生有意义的模型（即认知）不确定性（EU）表征尚不清楚。我们旨在调和这些冲突观点，指出两者均成立但源于不同的学习情境。值得注意的是，我们证明捷径的存在对EU表现为分歧具有决定性作用。
英文摘要
The correct way to quantify predictive uncertainty in neural networks remains a topic of active discussion. In particular, it is unclear whether the state-of-the art entropy decomposition leads to a meaningful representation of model, or epistemic, uncertainty (EU) in the light of a debate that pits ignorance against disagreement perspectives. We aim to reconcile the conflicting viewpoints by arguing that both are valid but arise from different learning situations. Notably, we show that the presence of shortcuts is decisive for EU manifesting as disagreement.
发表机构
- Department of Statistics, LMU Munich（慕尼黑大学统计系）
- Munich Center for Machine Learning (MCML)（慕尼黑机器学习中心）
机构由 AI 辅助整理，请以论文原文为准。
