发表机构
University of York; Cyprus University of Technology（约克大学; 塞浦路斯理工大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
CLEAR是一种轻量级事后方法，通过校准潜在空间几何并检测潜在冲突，在不重训练的情况下增强证据深度学习对分布外和对抗输入的鲁棒性，显著提升AUROC并保持性能。
AI 中文摘要
可靠的不确定性量化对于在高风险环境中部署深度学习模型至关重要，在这些环境中，分布外和对抗性输入可能导致模型产生自信但不可靠的预测。证据深度学习在单次前向传播中提供高效的不确定性估计，但仍可能对学习表示支持不足的输入（如对抗性输入）赋予较高的证据强度。我们提出了CLEAR，一种轻量级、任务无关的事后方法，无需重新训练或改变基础预测即可提高证据鲁棒性。利用留出的校准数据，CLEAR刻画了模型潜在空间中组条件几何结构。在推理时，它直接在潜在空间中高效生成扰动视图，并测量这些视图相对于预测组的校准几何结构的冲突程度。高潜在冲突表明证据缺乏支持，CLEAR利用这一点选择性地降低证据强度，同时保留对潜在一致输入的证据。在ImageNet→CUB上，CLEAR将OOD和对抗性AUROC分别提高了+8.29和+5.01，同时运行速度比竞争的事后方法快17.4倍，并在分类、回归和物体检测基准上保持预测性能。
英文摘要
Reliable uncertainty quantification is essential for deploying deep learning models in high-stakes settings, where out-of-distribution and adversarial inputs can induce confident but unreliable predictions. Evidential Deep Learning provides efficient uncertainty estimates in a single forward pass, but can still assign high evidential strength to inputs that are poorly supported by the learned representation, such as adversarial inputs. We introduce CLEAR, a lightweight, task-agnostic post-hoc method that improves evidential robustness without retraining or altering the base prediction. Using held-out calibration data, CLEAR characterises the group-conditioned geometry of the model's latent space. At inference, it efficiently generates perturbation views directly in the latent space and measures their conflict relative to the calibrated geometry of the predicted group. High latent conflict indicates unsupported evidence, which CLEAR uses to selectively reduce evidential strength while retaining evidence for latent-consistent inputs. On ImageNet$\rightarrow$CUB, CLEAR improves OOD and adversarial AUROC by $+8.29$ and $+5.01$ while running 17.4$\times$ faster than competing post-hoc methods while preserving predictive performance across classification, regression, and object detection benchmarks.
