AWhat changed
Demonstrated that few-shot exemplars with intermediate steps can improve several reported reasoning benchmarks for large language models.
Scientific scope
The record is the exact nine-author NeurIPS 2022 paper. Results are model-, prompt- and benchmark-specific; generated rationales are not proofs of faithful internal reasoning, factuality, safety or universal task improvement.
Evidence
The linked sources identify this contribution and its scope. WRS has not independently reconstructed the scientific claims. The record is the exact nine-author NeurIPS 2022 paper. Results are model-, prompt- and benchmark-specific; generated rationales are not proofs of faithful internal reasoning, factuality, safety or universal task improvement.
Contributors
- Jason Wei · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
- Xuezhi Wang · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
- Dale Schuurmans · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
- Maarten Bosma · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
- Brian Ichter · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
- Fei Xia · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
- Ed H. Chi · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
- Quoc V. Le · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
- Denny Zhou · SCIENCE: Named source author of the scoped work; individual effort shares are not measured.
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction