EvalLM: Interactive Evaluation of Large Language Model Prompts on User-Defined Criteria

要旨

By simply composing prompts, developers can prototype novel generative applications with Large Language Models (LLMs). To refine prototypes into products, however, developers must iteratively revise prompts by evaluating outputs to diagnose weaknesses. Formative interviews (N=8) revealed that developers invest significant effort in manually evaluating outputs as they assess context-specific and subjective criteria. We present EvalLM, an interactive system for iteratively refining prompts by evaluating multiple outputs on user-defined criteria. By describing criteria in natural language, users can employ the system's LLM-based evaluator to get an overview of where prompts excel or fail, and improve these based on the evaluator's feedback. A comparative study (N=12) showed that EvalLM, when compared to manual evaluation, helped participants compose more diverse criteria, examine twice as many outputs, and reach satisfactory prompts with 59% fewer revisions. Beyond prompts, our work can be extended to augment model evaluation and alignment in specific application contexts.

著者
Tae Soo Kim
KAIST, Daejeon, Korea, Republic of
Yoonjoo Lee
KAIST, Daejeon, Korea, Republic of
Jamin Shin
NAVER AI Lab, Seoul, Korea, Republic of
Young-Ho Kim
NAVER AI Lab, Seongnam, Gyeonggi, Korea, Republic of
Juho Kim
KAIST, Daejeon, Korea, Republic of
論文URL

https://doi.org/10.1145/3613904.3642216

動画

会議: CHI 2024

The ACM CHI Conference on Human Factors in Computing Systems (https://chi2024.acm.org/)

セッション: Evaluating AI Technologies B

320 'Emalani Theater
5 件の発表
2024-05-14 18:00:00
2024-05-14 19:20:00