PulseAugur
中
实时 23:58:32
English(EN) CIDER: A Dataset of Contextual Disclosure Boundaries for Privacy Preference Alignment

新的CIDER数据集助力大语言模型与用户隐私偏好对齐

作者 PulseAugur 编辑部 · [1 个来源] ·

研究人员推出CIDER，一个旨在帮助大语言模型更好地与个人隐私偏好对齐的新数据集。该数据集包含超过14,000条跨不同沟通场景的人工标注，详细说明了用户分享个人信息的意愿。实验表明，像GPT-5.4和Claude Sonnet 4.6这样的模型，在通过上下文信息进行个性化后，预测准确率可提高多达11.41个百分点，尽管这有时会导致不平衡的错误率。 AI

影响 该数据集有望促使大语言模型在实际互动中更好地尊重用户隐私。

排序理由 该集群描述了在arXiv上发布的一个新数据集和研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.AI 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

新的CIDER数据集助力大语言模型与用户隐私偏好对齐

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
该集群描述了在arXiv上发布的一个新数据集和研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, safety
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
58 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. arXiv cs.AI TIER_1 English(EN) · Bingcan Guo, Eryue Xu, Jijie Zhou, Zhiping Zhang, Tianshi Li ·

    CIDER：用于隐私偏好对齐的上下文披露边界数据集

    arXiv:2608.09164v1 Announce Type: new Abstract: Aligning large language models (LLMs) with human privacy preferences requires capturing individuals' disclosure boundaries beyond general privacy norms. However, a gap remains in eliciting such nuanced preferences to evaluate alignm…

相关实体

相关话题