Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Hallucination prediction on HotpotQA (original)
Loading...
83.39
AUROC
Conf + Probe (SCAO)
72.9276
75.6438
78.36
81.0762
Sep 18, 2025
AUROC
A(ϕ(s_M))
Updated 1mo ago
Evaluation Results
Method
Method
Links
AUROC
A(ϕ(s_M))
Conf + Probe (SCAO)
Model=LLaMA-8B
2025.09
83.39
15.02
Conf + Probe
Model=LLaMA-8B
2025.09
81.08
12.71
Probednn
Model=LLaMA-8B
2025.09
80.58
12.21
Conf (SCAO)
Model=LLaMA-8B
2025.09
77.7
-
Conf + Probe
Model=LLaMA-70B
2025.09
77.33
13.55
Conf + Probe (SCAO)
Model=LLaMA-70B
2025.09
77.28
13.5
Conf
Model=LLaMA-8B
2025.09
74.88
-
Probednn
Model=LLaMA-70B
2025.09
74.41
10.63
Conf (SCAO)
Model=LLaMA-70B
2025.09
74.13
-
Conf
Model=LLaMA-70B
2025.09
73.33
-
Feedback
Search any
task
Search any
task