Hugging Face Blog·· 2024-01-29AI 评分35
The Hallucinations Leaderboard：衡量大语言模型幻觉的开放基准
The Hallucinations Leaderboard, an Open Effort to Measure Hallucinations in Large Language Models
AI 导读
The Hallucinations Leaderboard 是一个开放项目，旨在通过特定基准评估大语言模型在事实性和忠实性方面的幻觉问题。该基准利用 EleutherAI Language Model Evaluation Harness，涵盖闭卷问答、摘要、阅读理解、指令遵循、事实核查及幻觉检测等任务。
来源：Hugging Face Blog · huggingface.co