Commonsense Reasoning on XCOPA

74.5Accuracy

Sailor2

Updated 1mo ago

Evaluation Results

Method	Links
Sailor2 2026.02		74.5
SeaLLM v3 2026.02		70.53
SEA-LION v3.5 2026.02		70.37
Multilingual 2026.02		70.3
Parallel Only (OPENSEAL) 2026.02		70.2
XGLM7.5B 2022.04		62.5
XGLM7.5B 2022.04		61.4
BYOL-mri 2026.01		61.2
1B Model (Multilingual classifier) 2026.04		60.8
XGLM7.5B 2022.04		60.6
BYOL-mri 2026.01		60.6
Gemma 2 2024.08		59.7
1B Model (HQ filtering) 2026.04		59.2
1B Model (No filtering) 2026.04		58.6
mGPT13B 2022.04		58.3
mGPT13B 2022.04		58.2
mGPT13B 2022.04		57.9
XGLM 2024.08		57.6
mGPT13B 2022.04		57.3
BYOL-mri 2026.01		57
XGLM1.7B 2022.04		56.8
XGLM1.7B 2022.04		56.2
XGLM1.7B 2022.04		56.1
LLaMA 3 2024.08		55.9
mGPT1.3B 2022.04		55.5
XGLM1.7B 2022.04		55.5
XGLM7.5B 2022.04		55.5
XGLM 2024.08		55.2
Goldfish 2024.08		55.1
mGPT1.3B 2022.04		54.9
mGPT1.3B 2022.04		54.8
mGPT1.3B 2022.04		54.8
Gemma 2 2024.08		54.6
BLOOM 2024.08		54.4
Chance 2024.08		50