← All organizations
Model evaluation
Vals AI
Vals AI builds benchmarks for evaluating language models on professional tasks. Its finance work examines both the accuracy of an answer and the resources needed to produce it.
Tax questions and leveraged-buyout spreadsheets illustrate why Vals evaluates individual tasks. A model that looks economical on a rate card can consume more tokens and deliver fewer usable answers on the work a firm actually assigns.
1 talk
Newest firstLoading more talks and speakers…
1 speaker at AIE
Affiliations reflect their AIE event listings and recorded appearances, not necessarily current employment.
