Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Hierarchical Instruction Following on IHEval Safety Defense Aligned setting
Loading...
63.2
Extract Accuracy (E)
Vanilla
54.36
56.655
58.95
61.245
Jun 29, 2026
Extract Accuracy (E)
Hijack Accuracy (H)
Average Accuracy
Updated 25d ago
Evaluation Results
Method
Method
Links
Extract Accuracy (E)
Hijack Accuracy (H)
Average Accuracy
Vanilla
Backbone=Llama-3.1-8B-...
2026.06
63.2
61.3
62.3
ISE
Backbone=Llama-3.1-8B-...
2026.06
62.6
61.1
61.9
IHDec
Backbone=Llama-3.1-8B-...
2026.06
54.7
53.9
54.3
Feedback
Search any
task
Search any
task
