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

Anthropic Harmless

Benchmarks

Task NameDataset NameSOTA ResultTrend
Reward ModelingAnthropic Harmless
Accuracy74.7
12
Reward Model TransferAnthropic Harmless (AHar)
AOG4.62
8
Open-ended DialogueAnthropic-Harmless ID
Win Rate68.1
4
Showing 3 of 3 rows