English(EN) Credal Concept Bottleneck Models for Epistemic-Aleatoric Uncertainty Decomposition

新的 CREDENCE 框架分解 AI 概念不确定性以改进决策

作者 PulseAugur 编辑部 · [2 个来源] · 2026-04-27 08:27

研究人员开发了 CREDENCE，一个用于 Credal 概念瓶颈模型 (CBM) 的新框架，该框架能有效分离预测中的认知不确定性和随机不确定性。这种分解使得更细致的决策成为可能，例如自动化低不确定性任务或将模糊案例转交人工审核。该框架将概念表示为概率区间，区分了可减少的模型欠指定和不可减少的输入歧义。 AI

影响通过区分模型局限性和固有的数据歧义，实现更精确的 AI 决策。

排序理由该集群描述了一篇详细介绍 AI 模型不确定性分解新框架的学术论文。

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。我们如何撰写摘要 →

报道来源 [2]

arXiv cs.AI TIER_1 English(EN) · Tanmoy Mukherjee, Thomas Bailleux, Pierre Marquis, Zied Bouraoui · 2026-04-28 04:00

Credal Concept Bottleneck Models for Epistemic-Aleatoric Uncertainty Decomposition

arXiv:2604.24170v1 Announce Type: new Abstract: Concept Bottleneck Models (CBMs) predict through human-interpretable concepts, but they typically output point concept probabilities that conflate epistemic uncertainty (reducible model underspecification) with aleatoric uncertainty…
Hugging Face Daily Papers TIER_1 English(EN) · 2026-04-27 08:27

Credal Concept Bottleneck Models for Epistemic-Aleatoric Uncertainty Decomposition

Concept Bottleneck Models (CBMs) predict through human-interpretable concepts, but they typically output point concept probabilities that conflate epistemic uncertainty (reducible model underspecification) with aleatoric uncertainty (irreducible input ambiguity). This makes conce…

报道来源 [2]

Credal Concept Bottleneck Models for Epistemic-Aleatoric Uncertainty Decomposition

Credal Concept Bottleneck Models for Epistemic-Aleatoric Uncertainty Decomposition

相关实体

相关话题