DROP
更新于 2026-04-03
801 次浏览
问题数量
96000
发布机构
Allen Institute for AI
评测类别
阅读理解
评测指标
F1
支持语言
英语
难度等级
Advanced
简介
一个需要模型进行离散推理的阅读理解基准，包括计数、比较和排序等操作。
相关资源
DROP Model Score Leaderboard
Source: DataLearnerAI
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
模式说明:
normal
thinking
low
medium
high
deeper thinking
parallel_thinking
图表加载中...
Latest DROP model rankings and full benchmark leaderboard
Browse the latest scores, model modes, release dates, and parameter sizes for DROP.
Model release cutoff:
DROP详细排名数据表格
| 排名 | 模型 | |||
|---|---|---|---|---|
| 1 | Pangu Pro MoE Standard Mode | 91.20 | 2025-06-30 | 719 |
| 2 | ERNIE-4.5-300B-A47B Standard Mode | 91.10 | 2025-06-30 | 3000 |
| 3 | DeepSeek-V3-0324 Standard Mode | 89.70 | 2025-03-24 | 6710 |
| 4 | GPT-4.1 Standard Mode | 89.20 | 2025-04-14 | 未知 |
| 5 | Qwen3-235B-A22B Standard Mode | 88.70 | 2025-04-28 | 2350 |
| 6 | Claude3-Opus Standard Mode | 83.10 | 2024-03-04 | 未知 |
| 7 | GPT-4 Standard Mode | 80.90 | 2023-03-14 | 1750 |
| 8 | Gemma 3 - 27B (IT) Standard Mode | 77.20 | 2025-03-12 | 270 |
| 9 | Gemma2-27B Standard Mode | 74.20 | 2024-05-14 | 270 |