SOTA Abductive Natural Language Inference on ANLI (leaderboard) and PapersWithCode

93.2Accuracy

UD+-XXL

Updated 4d ago

Evaluation Results

Method	Links
UD+-XXL 2022.11		93.2
Human Performance 2024.01		91.4
Tay et al. (2022) 2022.11		89.8
DeBERTa-v3-L (Supervised) 2024.01		89
UNICORN 2021.03		87.3
L2R2 2021.03		86.8
ROBERTa-L (Supervised) 2024.01		85.6
ROBERTA-LARGE 2021.03		83.9
VERA-T5 (Multitask Supervised) 2024.01		83.9
DeBERTa-v3-L (CANDLE Distilled) 2024.01		81.2
CAR-DEBERTa-v3-L 2024.01		79.6
CAR-DEBERTa-v3-L 2024.01		78.9
DeBERTa-v3-L (MR) 2024.01		76
DeBERTa-v3-L (MR) 2024.01		75.1
GPT-4 2024.01		75
VERA-T5-xxl (CANDLE Distilled) 2024.01		73.8
ChatGPT + Self-consistent chain-of-thought 2024.01		73.2
VERA-T5-xxl 2024.01		73.2
CAR-ROBERTa-L 2024.01		72.7
MKIF 2024.01		72.5
ZS-Fusion 2024.01		72.4
CAR-ROBERTa-L 2024.01		72.3
STL-PLM 2024.01		71.6
STL-Adapter 2024.01		71.5
STL-Adapter 2024.01		71.3
VERA-T5-xxl 2024.01		71.2
ROBERTa-L (MR) 2024.01		70.8
ROBERTa-L (MR) 2024.01		70.8
ROBERTa-L (MR) 2024.01		70.5
ChatGPT + Chain-of-thought 2024.01		70.5
VERA-T5-xxl 2024.01		70.3
ROBERTa-L (MR) 2024.01		70
MTL 2024.01		69.8
MTL 2024.01		69.6
ZS-Fusion 2024.01		69.6
ChatGPT 2024.01		69.3
BERT-LARGE 2021.03		66.8
ROBERTa-L 2024.01		65.5
SMLM 2024.01		65.3
GPT-3.5 2024.01		61.8
DeBERTa-v3-L 2024.01		59.9
LLAMA2 2024.01		57.5
GPT2-L 2024.01		56.5
LLAMA2 2024.01		55.9
Mistral-v0.1 2024.01		51
Majority Vote 2024.01		50.8
Random Vote 2024.01		50