Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

QASC

Benchmarks

Task NameDataset NameSOTA ResultTrend
Question AnsweringQASC
Score89.6
36
Multiple Choice Question AnsweringQASC
Accuracy100
22
Multiple Choice Question AnsweringQASC (test)
Accuracy78.5
21
Scientific Reasoning Question AnsweringQASC
Accuracy74.61
15
Question AnsweringQASC
Recall@174.17
15
Science Question AnsweringQASC (test)
Accuracy73.5
14
Commonsense ReasoningQASC (dev)
Accuracy84.02
14
Question AnsweringQASC
Cohen's d0.803
12
Question AnsweringQASC
Spearman's rho0.2727
12
Answer Plausibility EstimationQASC
Cohen's d0.803
10
Question AnsweringQASC
F114.73
10
Multiple Choice Question AnsweringQASC (dev)
Accuracy67.61
10
Chunking Strategy Evaluation for RAGQASC Evaluation Set (5-fold cross-validation)
Precision85
9
Question AnsweringQASC
Leakage Error14
9
Logical Refinement of Natural Language ExplanationsQASC
Initial Score17
8
Domain-specific Question Answeringqasc
Accuracy68.36
7
Commonsense Question AnsweringQASC (dev)
Accuracy83.7
7
Commonsense ReasoningQASC (test)
Accuracy90.06
6
Commonsense Question AnsweringScientific Commonsense (QASC) 1.0 (test)
Accuracy53.04
5
Question AnsweringQASC MRQA few-shot
F1 Score99.1
5
Commonsense Question AnsweringQASC
Accuracy72.8
4
Question AnsweringQASC
Accuracy43
2
Showing 22 of 22 rows