7261 次浏览 · 58 赞 · 6 转发 （收录时）
GLM-5.3 在 AA-PersonalityBench 中使用的破折号最多
- 谁
- Artificial Analysis
- 引文
In AA-PersonalityBench, GLM-5.3 uses the most em dashes of any model measured, averaging 7.4 per response, while Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 5.5 use almost none
发生了什么
在 AA-PersonalityBench 中，GLM-5.3 每条回复平均使用 7.4 个破折号，是所有受测模型中最多的。GLM-5.3-Flash 和 Muse Wonder 平均为 6.9 个，而 GPT-3.5 Turbo 和 Claude Haiku 5.5 均为 0.00 个。
接下来读什么
查看这件事的完整进展Artificial Analysis Compares Em Dash Use Across AA-PersonalityBench ModelsIn AA-PersonalityBench, GLM-5.3 uses the most em dashes of any model measured, averaging 7.4 per response, while Claude Opus 5.5, Claude Sonnet 5.5 and Claude Haiku 5.5 use almost none
We count the em dashes in every response each model gives across the benchmark. Claude Fable 5.1 averages 6.4 em dashes per response, while Claude Opus 5.5 averages fewer than one every 40 responses.
Average em dashes per response:
➤ Highest: GLM-5.3 (7.4), GLM-5.3-Flash (6.9), Muse Wonder (6.9), Kimi K3 (6.7) and Muse Spark 1.3 (6.5)
➤ Lowest: GPT-3.5 Turbo (0.00), Claude Haiku 5.5 (0.00), Claude Sonnet 5.5 (0.00 at medium, 0.01 at max), Llama 4 Maverick (0.01) and Claude Opus 5.5 (0.02)
这件事到目前为止
Artificial Analysis reported that GLM-5.3 averaged 7.4 em dashes per response in AA-PersonalityBench, the highest count among models measured. GLM-5.3-Flash and Muse Wonder averaged 6.9, while GPT-3.5 Turbo and Claude Haiku 5.5 averaged 0.00.
每天一封，跟上 AI 的重要变化
模型发布、价格变化、算力进展，附原始来源。免费，确认后开始发送，随时一键退订。 看一份往期日报 · 2026-11-04