Common Sense Reasoning on WG

94.1Accuracy

Human

Updated 29d ago

Evaluation Results

Method	Links
Human 2026.03		94.1
DeBERTa-v3-L (Supervised) 2026.03		84.1
IMAGINE-DeBERTa-v3-L 2026.03		79.3
RoBERTa-L (Supervised) 2026.03		79.3
IMAGINE-DeBERTa-v3-L (Retrieval) 2026.03		79.2
Qwen3-235B-A22B 2026.06		79.01
CANDLE-DeBERTa-v3-L 2026.03		78.3
CAR-DeBERTa-v3-L 2026.03		78.2
Llama2-70B 2026.06		77.58
GPT-4 (gpt-4) 2026.03		77
IMAGINE-DeBERTa-v3-L 2026.03		76.7
DeBERTa-v3-L (MR) 2026.03		76
CAT-Q 2026.06		75.34
Mistral-v0.1-7B 2026.03		75.3
Qwen3-32B 2026.06		73.56
Qwen3-14B 2026.06		73.16
LLAMA2-13B 2026.03		72.8
CAT-Q 2026.06		71.74
CANDLE-VERA-T5-xxl 2026.03		71.3
RF2-100B-A6.1B 2026.06		70.32
Qwen3-30B-A3B 2026.06		69.93
CAT-Q 2026.06		69.85
Llama2-7B 2026.06		69.46
VERA-T5-xxl 2026.03		68.1
Qwen3-8B 2026.06		68.03
CAT-Q 2026.06		68.03
VERA-T5-xxl 2026.03		67.2
Qwen3-4B 2026.06		65.51
CAT-Q 2026.06		65.43
ChatGPT (gpt-3.5-turbo) 2026.03		64.1
CAT-Q 2026.06		63.93
CAT-Q 2026.06		63.85
CAT-Q 2026.06		63.51
CAT-Q 2026.06		62.67
CAT-Q 2026.06		62.19
CAR-RoBERTa-L 2026.03		62
Qwen3-1.7B 2026.06		61.88
GenKnowSub 2025.05		61.24
IMAGINE-RoBERTa-L 2026.03		61.2
Multi-hop Knowledge Injection 2026.03		61
CAT-Q 2026.06		60.93
Arrow 2025.05		60.85
Zero-shot Fusion 2026.03		60.8
GPT-3.5 (text-davinci-003) 2026.03		60.7
GenKnowSub 2025.05		60.69
H128-MQA-CLA2 2024.05		60.69
H128-MQA 2024.05		60.46
RoBERTa-L (MR) 2026.03		59.6
InstructBLIP-Vicuna-7B 2026.03		59.6
H64-MQA 2024.05		57.85
RoBERTa-L 2026.03		57.5
Phi-2 2025.05		56.51
CAR-GPT-2-L 2026.03		55.2
IMAGINE-GPT-2-L 2026.03		55.2
CAT-Q 2026.06		54.83
Self-talk 2026.03		54.7
GPT-2-L (MR) 2026.03		54.7
CAT-Q 2026.06		54.62
LLaVA-1.5-7B 2026.03		54.5
GPT-2-L 2026.03		53.2
DeBERTa-v3-L 2026.03		50.3