top of page

Your certificate is now private

CertificateBackground.png

Certificate of Completion

ErdosHorizontal.png

THIS ACKNOWLEDGES THAT

HAS COMPLETED THE FALL 2026 DATA SCIENCE BOOT CAMP

Aprameya Girish Hebbar

Roman Holowinsky, PhD

November 15, 2026

DIRECTOR

DATE

clear.png

TEAM

The Right Answer, for Less

Guanhua Shao, Aprameya Girish Hebbar, Haotan Han, Garen Chiloyan, Eunice Ng, Han Cao

clear.png

Different language models (LLMs) offer different tradeoffs between performance and cost, so choosing the right model for each problem can significantly improve efficiency. For each new problem, we want to recommend the least expensive open language model that is likely to solve it correctly. We’ll learn from past model responses to predict both the chance of success and the dollar cost of input and output tokens before choosing a model. We’ll test whether this approach reduces spending while maintaining a chosen accuracy target. The goal is to decide when a cheaper model is sufficient and when paying for a stronger model is worthwhile.

Screen Shot 2022-06-03 at 11.31.35 AM.png
github URL

©2017-2026 by The Erdős Institute.

bottom of page