Share your thoughts, 1 month free Claude Pro on usSee more

Language Understanding on MMLU (Score)

88.6MMLU Score

gpt-oss-120b

Updated 12d ago

Evaluation Results

Method	Links
gpt-oss-120b 2026.01		88.6
Solar Open 2026.01		88.2
gpt-oss-120b 2026.01		87.9
GLM-4.5-Air 2026.01		83.3
Qwen2-72B-Instruct 2024.07		82.3
Llama-3-70B-Instruct 2024.07		82
Dense 2025.10		80.47
Dense 2025.10		76.82
Qwen1.5-110B-Chat 2024.07		76.5
Qwen1.5-72B-Chat 2024.07		75.6
Mixtral-8x22B-Instruct 2024.07		74
ARMOR 2025.10		71.43
SGDPO 2025.05		70.69
SamPO 2025.05		70.67
DPO 2025.05		70.65
BCO 2025.05		70.65
SFT 2025.05		70.6
NCA 2025.05		70.57
TDPO 2025.05		70.55
IPO 2025.05		70.39
System Prompt 2026.03		67
ARMOR 2025.10		66.22
Llama-3.1-8B-SFT 2026.03		66
VCL 2026.03		65
SparseGPT 2025.10		64.73
DRO 2026.03		64
Self-CD 2026.03		63
Wanda 2025.10		62.93
Vector Ablation 2026.03		62
NoWag-P 2025.10		61.69
NPO+KL 2025.05		61.4
Unilogit+KL 2025.05		61.4
Baseline 2025.05		60.8
ME+GD 2025.05		60.8
NPO+KL 2025.05		60.4
NPO 2025.05		60.2
UnDIAL+KL 2025.05		60.2
GA 2025.05		59.6
SimNPO+KL 2025.05		59.6
RKLD+KL 2025.05		59.5
SimNPO+KL 2025.05		59.1
GA+KL 2025.05		58.7
UnDIAL+KL 2025.05		58.5
SparseGPT 2025.10		55.77
Wanda 2025.10		55.75
Vanilla 2024.06		54.9
Unilogit+KL 2025.05		54.4
NoWag-P 2025.10		54.1
PaCE 2024.06		53.1
Qwen2-1.5B 2024.07		52.4
Prompting 2024.06		52.4
OrthoProj 2024.06		51.1
VecAdd 2024.06		50.9
Qwen1.5-1.8B 2024.07		43.7
Vanilla 2024.06		43.4
PaCE 2024.06		38.4
Qwen2-0.5B 2024.07		37.9
Qwen1.5-0.5B 2024.07		35
OrthoProj 2024.06		34.1
MoE 2026.02		33.24
Expert Divergence Learning 2026.02		33.21
Expert Divergence Learning 2026.02		32.98
MoE 2026.02		32.54
Expert Divergence Learning 2026.02		32.05
Expert Divergence Learning 2026.02		31.69
Expert Divergence Learning 2026.02		31.47
Expert Divergence Learning 2026.02		31.41
MoE 2026.02		31.16
VecAdd 2024.06		30.6
Prompting 2024.06		16.3