English(EN) Before Fine-Tuning: What LLMs Actually Are and How They Learn to Speak

新的大语言模型显示出比旧模型更低的句法和词汇多样性

作者 PulseAugur 编辑部 · [3 个来源] · 2026-05-03 11:51

一篇新发表在arXiv上的研究分析了大语言模型（LLMs）与人类书写文本相比的语法和词汇多样性。研究人员发现，与旧模型和《纽约时报》上的人类撰写的文章相比，新近的指令调优LLMs表现出更低的句法和词汇多样性。这表明，虽然指令调优提高了连贯性，但也可能缩小了LLM输出的表达范围。 AI

影响表明指令调优可能会缩小大语言模型的表达范围，影响创意和细致的文本生成。

排序理由分析大语言模型输出特征的学术论文。

AI 生成摘要 · Google Gemini · 来自 3 个来源。我们如何撰写摘要 →

报道来源 [3]

arXiv cs.CL TIER_1 English(EN) · Adri\'an Gude, Roi Santos-R\'ios, Francis Bond, Dan Flickinger, Carlos G\'omez-Rodr\'iguez, Olga Zamaraeva · 2026-05-08 04:00

More Aligned, Less Diverse? Analyzing the Grammar and Lexicon of Two Generations of LLMs

arXiv:2605.06030v1 Announce Type: new Abstract: This study contributes to a growing line of research in comparing LLM-generated texts with human-authored text, in this case, English news text. We focus in particular on the evaluation of syntactic properties through formal grammar…
arXiv cs.CL TIER_1 English(EN) · Olga Zamaraeva · 2026-05-07 11:21

More Aligned, Less Diverse? Analyzing the Grammar and Lexicon of Two Generations of LLMs

This study contributes to a growing line of research in comparing LLM-generated texts with human-authored text, in this case, English news text. We focus in particular on the evaluation of syntactic properties through formal grammar frameworks. Our analysis compares two generatio…
Medium — fine-tuning tag TIER_1 English(EN) · Karan_bhutani · 2026-05-03 11:51

Before Fine-Tuning: What LLMs Actually Are and How They Learn to Speak

<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@karanbhutani477/before-fine-tuning-what-llms-actually-are-and-how-they-learn-to-speak-43669987ab7d?source=rss------fine_tuning-5"><img src="https://cdn-images-1.medium.com/max/1672/1*OIutaf2hU…