Inadequacies of Large Language Model Benchmarks in the Era of Generative Artificial Intelligence

arXiv:2402.09880v1 Announce Type: cross
Abstract: The rapid rise in popularity of Large Language Models (LLMs) with emerging capabilities has spurred public curiosity to evaluate and compare different LLMs, leading many researchers to propose their LLM benchmarks. Noticing preliminary inadequacies in those benchmarks, we embarked on a study to critically assess 23 state-of-the-art LLM benchmarks, using our novel unified evaluation framework through the lenses of people, process, and technology, under the pillars of functionality and security. Our research uncovered significant limitations, …

abstract artificial artificial intelligence arxiv benchmarks capabilities cs.ai cs.cl cs.cy cs.hc curiosity generative generative artificial intelligence intelligence language language model language models large language large language model large language models llm llm benchmarks llms public researchers study type