Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Hierarchical Instruction Following on IHEval Safety Defense Conflict setting
Loading...
50.5
Extract Accuracy (E)
IHDec
27.412
33.406
39.4
45.394
Jun 29, 2026
Extract Accuracy (E)
Hijack Accuracy (H)
Average Accuracy
Updated 26d ago
Evaluation Results
Method
Method
Links
Extract Accuracy (E)
Hijack Accuracy (H)
Average Accuracy
IHDec
Backbone=Llama-3.1-8B-...
2026.06
50.5
57
53.8
Vanilla
Backbone=Llama-3.1-8B-...
2026.06
33.4
24.9
29.2
ISE
Backbone=Llama-3.1-8B-...
2026.06
28.3
28.9
28.6
Feedback
Search any
task
Search any
task