大模型
x.com ·
@lmarena.ai(@lmarena_ai)
LMArena 对比 Claude Opus 5.5 与 Opus 5 的写作风格变化
LMArena 分析了 Claude Opus 5.5 与 Opus 5 在高推理 Text Arena 输出中的写作差异：12 项写作指标中 10 项改善，长实词占比从 41.7% 降至 38.6%，平均句长从 12.14 词缩短至 10.03 词（约 17%）；代价是回答更长，平均从 453 词增至 481 词，为 Opus 系列最长。破折号减少 95%、分号减少 73%，但“perhaps”“arguably”等模糊限定词上升 97%（每千词 0.39 升至 0.77），为所分析 Claude 模型中最高。
AI推荐理由 LMArena 对 Claude Opus 5.5 与 Opus 5 在高推理文本输出上的写作特征做了量化对比，给出长词占比、句长、回答长度、破折号/分号使用及模糊限定词频率等具体数值，属于可复核的模型行为分析，对关注模型输出风格与评测的开发者有参考价值。局限是样本仅限 Text Arena 高推理输出，指标为风格统计而非能力评测，且结论由平台自述、未提供完整方法细节。
72
浏览 1,050,242
点赞 1,609
评论 76
转发 96
正文
We analyzed how @claudeai’s Opus 5.5 writes compared with Opus 5 across high-reasoning Text Arena outputs.
10 of 12 writing measures moved in a better direction.
Opus 5.5 should be easier to read:
- Long content words fall from 41.7% to 38.6%, the lowest share of any Claude model we analyzed.
- Sentences are 17% shorter on average, dropping from 12.14 to 10.03 words.
The tradeoff is length. Answers get 6% wordier, rising from 453 to 481 words on average, making Opus 5.5 give the longest answers across the Opus family.
It also sounds less recognizably AI on two familiar tells:
- 95% fewer em dashes
- 73% fewer semicolons
But a new giveaway may be emerging. Hedges and caveats such as “perhaps” and “arguably” rise 97%, from 0.39 to 0.77 per 1,000 words, the highest rate of any Claude model we analyzed.
What do you think: does Opus 5.5 read more naturally?