List your company

Safety Evaluation on SORRY-Bench (expected HFR)

0.9Expected HFRG
-0.72410.23821.232.162Oct 10, 2025

Evaluation results

MethodLinks
GBase Model=GPT-OSS-20B2025.100.9
MIBase Model=GPT-OSS-20B2025.100.9
SGBase Model=LLaMA-2-7B2025.105.1
SCBase Model=GPT-OSS-20B2025.1011.3
SGBase Model=LLaMA-2-13B2025.1014.2
SGBase Model=Llama-3.1-8B2025.1016.7
ER-GBase Model=GPT-OSS-20B2025.1020.7
ER-GBase Model=Llama-3.2-1B2025.1022
SGBase Model=GPT-OSS-20B2025.1022.3
ER-DBase Model=Llama-3.2-1B2025.1024.5
DBase Model=Llama-3.2-1B2025.1026.3
SCBase Model=LLaMA-2-7B2025.1027.2
MIBase Model=Llama-3.2-1B2025.1027.3
GBase Model=Llama-3.2-1B2025.1027.6
SCBase Model=LLaMA-2-13B2025.1029.5
ER-DBase Model=GPT-OSS-20B2025.1030.1
ER-GBase Model=Llama-3.1-8B2025.1030.4
MIBase Model=LLaMA-2-7B2025.1030.7
GBase Model=LLaMA-2-7B2025.1030.8
ER-DBase Model=Llama-3.1-8B2025.1030.9
DBase Model=GPT-OSS-20B2025.1031
ER-GBase Model=LLaMA-2-7B2025.1031.2
SCBase Model=Llama-3.2-1B2025.1031.2
DBase Model=LLaMA-2-7B2025.1031.6
ER-DBase Model=LLaMA-2-7B2025.1031.6
SCBase Model=Llama-3.1-8B2025.1032.3
DBase Model=LLaMA-2-13B2025.1032.8
GBase Model=Llama-3.1-8B2025.1032.9
MIBase Model=Llama-3.1-8B2025.1032.9
DBase Model=Llama-3.1-8B2025.1033.2
GBase Model=LLaMA-2-13B2025.1033.5
MIBase Model=LLaMA-2-13B2025.1033.5
ER-DBase Model=LLaMA-2-13B2025.1033.5
ER-GBase Model=LLaMA-2-13B2025.1035.6
SGBase Model=Llama-3.2-1B2025.1041.5