Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Hierarchical Instruction Following on IHEval Safety Defense Reference setting
Loading...
78.9
Extract Accuracy
Vanilla
45.204
53.952
62.7
71.448
Jun 29, 2026
Extract Accuracy
Hijack Accuracy
Average Accuracy
Updated 26d ago
Evaluation Results
Method
Method
Links
Extract Accuracy
Hijack Accuracy
Average Accuracy
Vanilla
Backbone=Llama-3.1-8B-...
2026.06
78.9
79.1
79
IHDec
Backbone=Llama-3.1-8B-...
2026.06
66.7
78.1
72.4
ISE
Backbone=Llama-3.1-8B-...
2026.06
46.5
44.9
45.7
Feedback
Search any
task
Search any
task