Commonsense Question Answering on CommonsenseQA (CSQA) (val)

75.7Accuracy

ChatGPT + Self-consistent chain-of-thought

Updated 1mo ago

Evaluation Results

Method	Links
ChatGPT + Self-consistent chain-of-thought 2024.01		75.7
ChatGPT + Chain-of-thought 2024.01		75.5
ChatGPT 2024.01		74.5
DeBERTa-v3-L (MR) 2024.01		71.6
DeBERTa-v3-L (CANDLE Distilled) 2024.01		69.9
CAR-DeBERTa-v3-L 2024.01		69.3
GPT 3.5 2024.01		68.9
LLAMA2 2024.01		67.3
CAR-DeBERTa-v3-L 2024.01		67.2
DeBERTa-v3-L (MR) 2024.01		67
STL-Adapter 2024.01		66.5
VERA-T5-xxl (CANDLE Distilled) 2024.01		64.7
VERA-T5-xxl 2024.01		63
VERA-T5-xxl 2024.01		61.7
Mistral-v0.1 2024.01		59.6
VERA-T5-xxl 2024.01		59.5
LLAMA2 2024.01		57.8
ROBERTa-L 2024.01		45
MICO 2024.01		44.2
GPT-4 2024.01		43
SMLM 2024.01		38.8
Self-talk 2024.01		32.4
DeBERTa-v3-L 2024.01		25.4