Multi-turn benchmark with 80 challenging questions across 8 domains (writing, reasoning, math, coding, STEM, humanities, roleplay, extraction). Uses GPT-4-as-judge to score responses, enabling automated preference evaluation. Powers the Chatbot Arena leaderboard methodology.
| Provider | lmsys |
| Category | 评估与基准测试 |
| Size | 80 Questions |
| License | Apache 2.0 |
| Downloads | 900k |
| Tags | Benchmark, Multi-Turn, LLM-Judge, GPT-4, Chatbot-Arena |
from datasets import load_dataset
ds = load_dataset("lmsys/mt_bench_human_judgments")