multistep-reasoning

Tag

Cards List
#multistep-reasoning

Solving math word problems

OpenAI Blog · 2021-10-29 Cached

OpenAI trained a system using verifiers to solve grade school math word problems with 90% of child-level accuracy, nearly doubling fine-tuned GPT-3 performance. The approach addresses language models' weakness in multistep reasoning by training verifiers to evaluate candidate solutions and select the best one.

0 favorites 0 likes
← Back to home

Submit Feedback