NAACL (2025)
Seungone Kim, Juyoung Suk (KAIST), Ji Yong Cho, Shayne Longpre (MIT), Chaeeun Kim (KAIST), Dongkeun Yoon (KAIST), Guijin Son (Yonsei University), Yejin Cho (KAIST), Sheikh Shafayat (KAIST), Jinheon Baek (KAIST), Sue Hyun Park (KAIST), Hyeonbin Hwang (KAIST), Jinkyung Jo (KAIST), Hyowon Cho (KAIST), Haebin Shin (KAIST)
Abstract
As language models (LMs) become capable of handling a wide range of tasks, their evaluation is becoming as challenging as their development. Most generation benchmarks currently assess LMs using abstract evaluation criteria like helpfulness and harmlessness, which often lack the flexibility and granularity of human assessment. Additionally, these benchmarks tend to focus disproportionately on specific capabilities such as instruction following, leading to coverage bias. To overcome these limitations, we introduce the BiGGen Bench, a principled generation benchmark designed to thoroughly evaluate nine distinct capabilities of LMs across 77 diverse tasks. A key feature of the BiGGen Bench is its use of instance-specific evaluation criteria, closely mirroring the nuanced discernment of human evaluation. We apply this benchmark to assess 103 frontier LMs using five evaluator LMs. Our code, data, and evaluation results are all publicly available at this https URL.