两篇新研究论文指出了当前评估AI生成文本方法的重大问题。一篇论文揭示了NLP会议上普遍存在的人工评估协议报告不足的问题，阻碍了可复现性和清晰度。第二篇论文批评了非自回归模型常用生成式困惑度的方法，认为它可以被“破解”以生成不连贯的文本，同时表现良好。两项研究都呼吁采用更健壮和透明的评估指标及方法。 AI
影响 强调了当前AI文本评估中的关键缺陷，可能导致更可靠的基准和模型开发。
排序理由 两篇学术论文发表在arXiv上，讨论了AI文本评估指标和协议的基本问题。
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
arXiv:2606.07936v1 Announce Type: cross Abstract: Human evaluation plays a critical role in assessing the quality of generated text. However, the reliability and reproducibility of these evaluations depend on transparent and well-documented protocols -- details that are frequentl…
arXiv:2606.08417v1 Announce Type: cross Abstract: Diffusion and continuous flow-based language models have emerged as the leading non-autoregressive alternatives to language modeling. Progress in both paradigms is overwhelmingly tracked by generative perplexity (gen-PPL): the per…
Diffusion and continuous flow-based language models have emerged as the leading non-autoregressive alternatives to language modeling. Progress in both paradigms is overwhelmingly tracked by generative perplexity (gen-PPL): the per-token negative log-likelihood of samples under a …
Human evaluation plays a critical role in assessing the quality of generated text. However, the reliability and reproducibility of these evaluations depend on transparent and well-documented protocols -- details that are frequently missing in current practice. In this work, we co…