List your company
ResearchFormerly WizWand

World Knowledge on MMLU

82.4AccuracyLlama-3.3-70b-Instruct
22.39237.97153.5569.129Oct 10, 2023Mar 17, 2024Aug 24, 2024Jan 30, 2025Jul 9, 2025Dec 15, 2025May 24, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Llama-3.3-70b-InstructSize=70B, Few-shot settings=5-shot2026.0382.4
KarnakSize=40B, Few-shot settings=5-shot2026.0382.37
Qwen3-32BSize=32B, Few-shot settings=5-shot2026.0382.25
Fanar-27BSize=27B, Few-shot settings=5-shot2026.0378.89
AceGPT-v2-70B-ChatSize=70B, Few-shot settings=5-shot2026.0377.98
Gemma-3-27B-itSize=27B, Few-shot settings=5-shot2026.0377.38
AceGPT-v2-32B-ChatSize=32B, Few-shot settings=5-shot2026.0375.72
Jais-2-70B-ChatSize=70B, Few-shot settings=5-shot2026.0373.86
Fanar-1-9b-instructSize=9B, Few-shot settings=5-shot2026.0371.32
Allam-7B-Instruct-preview-v2Size=7B, Few-shot settings=5-shot2026.0363.8
ID-LoRABase Model=Mistral-7B, Trainable Parameters (%)=0.62%2026.0260.6
HydraLoRABase Model=Mistral-7B, Trainable Parameters (%)=1.22%2026.0260.4
MoELoRABase Model=Mistral-7B, Trainable Parameters (%)=1.21%2026.0260.4
DoRABase Model=Mistral-7B, Trainable Parameters (%)=1.16%2026.0259.4
LoRABase Model=Mistral-7B, Trainable Parameters (%)=1.15%2026.0259.3
ID-LoRABase Model=LLaMA-3-8B, Trainable Parameters (%)=0.56%2026.0251
LoRABase Model=LLaMA-3-8B, Trainable Parameters (%)=1.03%2026.0248.1
HydraLoRABase Model=LLaMA-3-8B, Trainable Parameters (%)=1.10%2026.0247.5
FFTBase Model=Mistral-7B, Trainable Parameters (%)=100%2026.0246.7
LLaMA2-7B# tokens for training=2T, Shot number=5, Training data source=Different from RedPajama2023.1046.6
NITPModel scale=9bA1b, Evaluation=few-shot, Context length=81922026.0546.14
NTPModel scale=9bA1b, Evaluation=few-shot, Context length=81922026.0543.71
DoRABase Model=LLaMA-3-8B, Trainable Parameters (%)=1.05%2026.0241.5
MoELoRABase Model=LLaMA-3-8B, Trainable Parameters (%)=1.09%2026.0241.5
FFTBase Model=LLaMA-3-8B, Trainable Parameters (%)=100%2026.0237.5
NITPModel scale=3bA0.5b, Evaluation=few-shot, Context length=81922026.0537.37
NTPModel scale=3bA0.5b, Evaluation=few-shot, Context length=81922026.0534.6
NITPModel scale=1.9bA0.3b, Evaluation=few-shot, Context length=81922026.0531.68
NTPModel scale=1.9bA0.3b, Evaluation=few-shot, Context length=81922026.0531.05
INCITE-Base-3B# tokens for training=800B, Shot number=5, Training data source=RedPajama2023.1027
Open-LLaMA-3B-v1# tokens for training=1T, Shot number=5, Training data source=RedPajama2023.1027
Pythia-2.8B# tokens for training=300B, Shot number=5, Training data source=RedPajama2023.1026.9
Open-LLaMA-3B-v2# tokens for training=1T, Shot number=5, Training data source=RedPajama2023.1026.9
Sheared-LLaMA-2.7B# tokens for training=50B, Shot number=5, Training data source=RedPajama2023.1026.4
OPT-2.7B# tokens for training=300B, Shot number=5, Training data source=Different from RedPajama2023.1025.9
Pythia-1.4B# tokens for training=300B, Shot number=5, Training data source=Different from RedPajama2023.1025.7
Sheared-LLaMA-1.3B# tokens for training=50B, Shot number=5, Training data source=RedPajama2023.1025.7
TinyLlama-1.1B# tokens for training=3T, Shot number=5, Training data source=RedPajama2023.1025.5
OPT-1.3B# tokens for training=300B, Shot number=5, Training data source=Different from RedPajama2023.1024.7