@Arthur63DeepSeek R1 LiveBench 部分分数 中发帖

Language




model
connections
plot_unscrambling
typos




deepseek-reasoner
74.167
43.046
44.0







model
average
language




deepseek-reasoner
53.7
53.7



Data Analysis




model
cta
tablejoin
tablereformat




deepseek-reasoner
64.0
61.46
90.0







model
average
data_analysis




deepseek-reasoner
71.8
71.8



Instruction Following




model
paraphrase
simplify
story_generation
summarize

...