Our new X account is live! Follow @wizwand_team for updates

Commonsense Reasoning on CommonsenseQA (CSQA) v1.0 (test)

64.11Accuracy

CKT-base

Updated 3d ago

Evaluation Results

Method	Links
CKT-base 2023.06		64.11
CALM 2023.06		63.32
T5-base + CSKG (Rule) 2023.06		63.1
T5-base + SSM 2023.06		62.37
T5-base + TI 2023.06		62.05
T5-base 2023.06		61.88
T5-base + KD 2023.06		61.83
CKT w/ GPT-2 2023.06		60.39
T5-base + CSKG (TI) 2023.06		60.22
ERNIE-base 2023.06		54.06
KnowBERT 2023.06		53.88
BERT-base 2023.06		53.08
COMET 2023.06		45.32
TituLLM-3b-v2.0 2025.02		33
BLOOM-3b 2025.02		32
BLOOM-1b7 2025.02		30
Llama-3.2-3b 2025.02		29
Gemma-2-2b 2025.02		28
TituLLM-1b-v2.0 2025.02		28
TituLLM-3b-v2.0 2025.02		28
BLOOM-1b1 2025.02		27
BLOOM-1b1 2025.02		27
BLOOM-1b7 2025.02		27
BLOOM-3b 2025.02		27
Llama-3.2-3b 2025.02		26
Gemma-2-2b 2025.02		26
BLOOM-560m 2025.02		26
BLOOM-560m 2025.02		26
TituLLM-1b-v2.0 2025.02		26
BongLLaMA-3.2-1b 2025.02		24
Llama-3.2-1b 2025.02		23
Llama-3.2-1b 2025.02		23
Qwen-2.5-1.5b 2025.02		23
Qwen-2.5-1.5b 2025.02		23
davinci 2025.02		22
Qwen-2.5-0.5b 2025.02		22
SmolLM2-135m 2025.02		22
SmolLM2-1.7b 2025.02		22
BongLLaMA-3.2-1b 2025.02		22
Qwen-2.5-0.5b 2025.02		21
SmolLM2-135m 2025.02		21
SmolLM2-360m 2025.02		21
SmolLM2-1.7b 2025.02		21
BongLLaMA-3.2-3b 2025.02		21
SmolLM2-360m 2025.02		20
BongLLaMA-3.2-3b 2025.02		20