Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Arena-Hard-Auto

Benchmarks

Task NameDataset NameSOTA ResultTrend
Open-ended query evaluationArena-Hard-Auto v0.1
Win Rate31.5
14
Chat PerformanceArena-Hard-Auto
Score92.8
6
Open-ended instruction followingArena-Hard-Auto v2.0
Win Rate (Response)51.8
3
Showing 3 of 3 rows