Common Sense Reasoning
The goal of Common Sense Reasoning tasks is to enable models to go beyond pattern recognition and utilize common sense or world knowledge for reasoning. This aims to allow models to understand complex situations, make reasonable judgments and predictions, thereby enhancing the intelligence level and application effects in fields such as natural language processing and dialogue systems.
ARC (Challenge)
ARC (Easy)
GAL 120B (0-shot)
BIG-bench (Causal Judgment)
BIG-bench (Date Understanding)
BIG-bench (Disambiguation QA)
BIG-bench (Known Unknowns)
PaLM-540B (few-shot, k=5)
BIG-bench (Logical Sequence)
Chinchilla-70B (few-shot, k=5)
BIG-bench (Sports Understanding)
BIG-bench (Winowhy)
CODAH
BERT Large
CommonsenseQA
QA-GNN
CrowdSource QA
Event2Mind dev
Event2Mind test
EA-VQ-VAE
PARus
ReCoRD
ST-MoE-32B 269B (fine-tuned)
RuCoS
Russian Event2Mind
araneum word2vec (skipgram) + GRU
RWSD
SWAG
DeBERTalarge
Visual Dialog v0.9
Visual Dialog v0.9
NMN [kottur2018visual]
WinoGAViL
ViLT
WinoGrande
PaLM 540B (0-shot)