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

FollowBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Instruction FollowingFollowBench
HSR79
85
Instruction FollowingFollowBench (test)
CSR85.1
24
Instruction FollowingFollowBench In-Domain
HSR70.4
23
Instruction FollowingFollowBench
SSR70.2
22
Constraint FollowingFollowBench SSR
Success Rate (Level 1)84.7
18
Reward ModelingFollowBench
Accuracy89.9
17
Instruction Following EvaluationFollowBench OOD
HSR78.06
14
Instruction FollowingFollowBench
CSR83.8
3
Showing 8 of 8 rows