Aligned Pāli Vinaya Explicit-Content Translation Dataset: Brahmali and Four Large Language Models (445 Segments)
收藏官方服务：
资源简介：
Complete 445-segment aligned analysis dataset accompanying the paper "Implicit Gatekeeping in AI Translation: Lexical Sanitization of Explicit Content in the Pāli Vinaya Piṭaka". Each record contains the Pāli source, Bhikkhu Brahmali’s English reference translation, outputs from GPT-5.2, Claude Opus 4.5, Grok 4 Fast, and Gemini 3 Pro Preview, and Claude-derived explicit-content screening scores and labels.
提供机构：
Zenodo创建时间：
2026-08-11
