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

CQA

Benchmarks

Task NameDataset NameSOTA ResultTrend
Question AnsweringCQA
Accuracy83.1
25
ReasoningCQA
Accuracy67.32
12
Information RetrievalcQA out-of-domain (test)
MAP@10052.4
8
Commonsense ReasoningCQA (evaluation)
Accuracy79.2
8
Information RetrievalcQA Scifi
MAP@10054.1
7
Information RetrievalcQA Gaming
MAP@10051.5
7
Information RetrievalcQA English
MAP@10054.3
7
Information RetrievalcQA Apple
MAP@10030.7
7
Community Question AnsweringcQA Scifi domain StackExchange (test)
MAP@10064.1
7
Community Question AnsweringcQA Gaming domain StackExchange (test)
MAP@1000.592
7
Community Question AnsweringcQA English domain StackExchange (test)
MAP@1000.606
7
Community Question AnsweringcQA Apple domain StackExchange (test)
MAP@10037.8
7
Chain-of-Thought GenerationCQA (test)
GPT-4 Score4.11
6
Question AnsweringCQA
Accuracy (GPT-2-Small)43
4
Commonsense Question AnsweringCQA (test)
Accuracy43.9
3
Commonsense Question AnsweringCQA
ECE11.75
2
Showing 16 of 16 rows