Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Span Extraction on Causal and Downstream Robustness Ablation Suite
Loading...
81
Span F1
HETA
36.28
47.89
59.5
71.11
Apr 14, 2026
Span F1
Updated 2d ago
Evaluation Results
Method
Method
Links
Span F1
HETA
Method Variant=Full
2026.04
81
HETA
Method Variant=LR+WIN
2026.04
78
HETA
Method Variant=w/o Hes...
2026.04
72
HETA
Method Variant=w/o KL
2026.04
69
HETA
Method Variant=w/o Tra...
2026.04
64
ReAGent
Method Variant=Standard
2026.04
63
SEA-CoT
Method Variant=Standard
2026.04
60
Progressive Inference
Method Variant=Standard
2026.04
58
fAML
Method Variant=Standard
2026.04
56
ContextCite
Method Variant=Standard
2026.04
55
TDD-backward
Method Variant=Standard
2026.04
54
Peering (PML)
Method Variant=Standard
2026.04
52
Integrated Gradients
Method Variant=Standard
2026.04
49
Attention Rollout
Method Variant=Standard
2026.04
38
Feedback
Search any
task
Search any
task