CALLIOPE：无需嵌入向量、支持差分隐私的 AI 使用洞察框架

CALLIOPE: Private Insights into AI Use, Without Embeddings

AI 导读

CALLIOPE 是一个自上而下发现大规模非结构化数据（如 LLM 对话）主题层级、并支持差分隐私（DP）的框架，整个流程直接调用 LLM 能力，不依赖向量嵌入聚类。实验显示其在 DP 设置下效用优于现有基于嵌入的方法，高隐私强度场景尤为明显。团队用传统聚类指标及新提出的效用指标评估了输出质量。

正文 · 原文

Authors:Edith Cohen, Vadym Doroshenko, Badih Ghazi, Charlie Harrison, Pritish Kamath, Alexander Knop, Ravi Kumar, Ethan Leeman, Pasin Manurangsi, Mikhail Pravilov, Adam Sealfon, Yurii Sushko, Marika Swanberg, Chiyuan Zhang

View PDF HTML (experimental)

Abstract:We introduce CALLIOPE, a top-down framework for discovering topic hierarchies in large corpora of unstructured data (e.g., LLM conversations), while supporting differential privacy (DP). Unlike traditional and even recent LLM-augmented approaches that ultimately rely on clustering vector embeddings, our framework leverages LLM capabilities directly for the entire pipeline. Experiments demonstrate that this approach provides better utility over existing (embedding-based) methods in the DP setting, especially in the high privacy regime. We evaluate the quality of CALLIOPE outputs using a variety of traditionally studied clustering metrics, as well as some novel utility metrics that we introduce.

Subjects:	Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Cryptography and Security (cs.CR)
Cite as:	arXiv:2610.28797 [cs.LG]
	(or arXiv:2610.28797v1 [cs.LG] for this version)
	https://doi.org/10.48550/arXiv.2610.28797 arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Ethan Leeman [view email]
[v1] Tue, 20 Oct 2026 18:23:08 UTC (709 KB)

来源：arXiv cs.CL · arxiv.org