← Back to feed

BudHist Benchmark: Testing the Knowledge of Buddhist History in Open-Weight LLMs

AI · Research · 2026-11-03 11:38

该项目用四选一题库评测本地开源权重模型对佛教历史的掌握程度，强调实验可复现与模型可归档，本地用 Ollama、云端用 OpenRouter。

Source: mbingenheimer.net

Auto-generated summary for reference only. RSS