arXiv cs.AI· Mohammad Jalali, Azim Ospanov, Amin Gohari, Farzan Farnia·· 6 天前AI 评分22
Conditional-Vendi：面向生成式 AI 模型与 LLM 的提示词感知多样性评估
Conditional Vendi Score: Prompt-Aware Diversity Evaluation for Generative AI Models and LLMs
AI 导读
研究者提出 Conditional-Vendi 与 Conditional-RKE 两项多样性指标，基于半正定矩阵的条件熵，用于在提示词引导的生成中剥离提示词带来的变异、单独衡量模型自身的多样性。
正文 · 原文
Abstract:Generative models guided by text prompts are widely evaluated for fidelity and prompt alignment, yet their ability to produce diverse outputs remains underexplored. Existing diversity metrics such as Vendi and RKE, which are based on the von Neumann and Rényi entropies of kernel matrices, were developed for unconditional models and cannot distinguish prompt-induced from model-induced variability. We address this gap by introducing \textit{Conditional-Vendi} and \textit{Conditional-RKE}, diversity measures derived from the conditional entropy of positive semidefinite matrices. These scores isolate model-induced diversity in prompt-guided generation, with Conditional-RKE enjoying an $O(1/\sqrt{n})$ convergence rate. For Conditional-Vendi, we introduce a truncated-spectrum approximation that yields scalable and consistent estimates. Experiments on text-to-image, image-captioning, and LLM tasks show that the conditional scores recover ground-truth diversity orderings and can also guide diffusion models toward more diverse samples. The codebase is available at this https URL.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Information Theory (cs.IT) |
| MSC classes: | 68T07 |
| Cite as: | arXiv:2411.02817 [cs.LG] |
| (or arXiv:2411.02817v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2411.02817 arXiv-issued DOI via DataCite |
Submission history
From: Mohammad Jalali [view email]
[v1]
Tue, 5 Nov 2024 05:30:39 UTC (5,111 KB)
[v2]
Tue, 9 Jun 2026 06:28:35 UTC (9,054 KB)
[v3]
Fri, 9 Oct 2026 09:58:18 UTC (9,054 KB)
来源：arXiv cs.AI · arxiv.org