Designing a strong test for measuring true common-sense reasoning

Santos, Henrique; McGuinness, Deborah; Kejriwal, Mayank; Mulvehill, Alice

Author

Santos, Henrique; McGuinness, Deborah; Kejriwal, Mayank; Mulvehill, Alice

Other Contributors

Date Issued

2022-04-22

Subject

Common-sense reasoning; artificial intelligence

Degree

Terms of Use

This Item is protected by copyright and/or related rights. ;

Full Citation

Kejriwal, M., Santos, H., Mulvehill, A.M. et al. Designing a strong test for measuring true common-sense reasoning. Nat Mach Intell 4, 318–322 (2022).

Metadata

Show full item record

URI

https://rdcu.be/cLUHR; https://hdl.handle.net/20.500.13015/4979

Abstract

Common-sense reasoning has recently emerged as an important test for artificial general intelligence, especially given the much-publicized successes of language representation models such as T5, BERT and GPT-3. Currently, typical benchmarks involve question answering tasks, but to test the full complexity of common-sense reasoning, more comprehensive evaluation methods that are grounded in theory should be developed.;