GPT-5.2 · MMLU Language (0-shot)(OpenAI GPT-5 system card, 2025-08-07; MMLU test set translated by professional human translators; chain-of-thought prompting. Section 3.10 says 13 languages, but Table 10 lists 14 language rows.) 89.5% · 附件证据(ev-gpt52-mmlu-language-korean)
GPT-5.2 的 MMLU Language (0-shot) · OpenAI GPT-5 system card, 2025-08-07; MMLU test set translated by professional human translators; chain-of-thought prompting. Section 3.10 says 13 languages, but Table 10 lists 14 language rows. 官方出处。GPT-5.2 Thinking (system card) · 89.5%。见《GPT-5.2 System Card Update PDF | OpenAI》PDF p. 11, Table 8 MMLU Language (0-shot, higher is better), row Korean, gpt-5.2-thinking column。证据 ev-gpt52-mmlu-language-korean。
原文与定位
GPT-5.2 · MMLU Language (0-shot)(OpenAI GPT-5 system card, 2025-08-07; MMLU test set translated by professional human translators; chain-of-thought prompting. Section 3.10 says 13 languages, but Table 10 lists 14 language rows.),GPT-5.2 Thinking (system card)配置,Accuracy 89.5%(成绩公布 2025-12-11)。见《GPT-5.2 System Card Update PDF | OpenAI》PDF p. 11, Table 8 MMLU Language (0-shot, higher is better), row Korean, gpt-5.2-thinking column。 查看成绩条件
Korean 0.896 0.895
- 官方材料
- GPT-5.2 System Card Update PDF | OpenAI · PDF 第 11 页
- 定位
- 附件 · PDF p. 11, Table 8 MMLU Language (0-shot, higher is better), row Korean, gpt-5.2-thinking column
- 材料版本
- 官方 27 页 PDF,完整核读 PDF p. 1-27,含图表页均已实际渲染检查;封面标注 December 11, 2025。
- 版本公布
- 披露时间
- · 已有首发依据
- 采集时间
- 历史依据
- 暂无可靠历史版本依据
引用文本
GPT-5.2 · MMLU Language (0-shot)(OpenAI GPT-5 system card, 2025-08-07; MMLU test set translated by professional human translators; chain-of-thought prompting. Section 3.10 says 13 languages, but Table 10 lists 14 language rows.),GPT-5.2 Thinking (system card)配置,Accuracy 89.5%(成绩公布 2025-12-11)。官方出处 GPT-5.2 System Card Update PDF | OpenAI(附件,官方 27 页 PDF,完整核读 PDF p. 1-27,含图表页均已实际渲染检查;封面标注 December 11, 2025。)PDF p. 11, Table 8 MMLU Language (0-shot, higher is better), row Korean, gpt-5.2-thinking column,https://cdn.openai.com/pdf/3a4153c8-c748-4b71-8e31-aecbde944f8d/oai_5_2_system-card.pdf#page=11。Benchlog https://benchlog.lekshome.cc/evidence/gpt-5-2/ev-gpt52-mmlu-language-korean/
关联成绩与条件
| 指标与成绩 | 配置与评测条件 | 证据关系 |
|---|---|---|
| Accuracy89.5%原文 0.895 · 单位 % 公布 | GPT-5.2 Thinking (system card)
| 直接支持该成绩 在排名中查看此成绩 |