Bias in Language Models: Beyond Trick Tests and Toward RUTEd Evaluation

arXiv:2402.12649v1 Announce Type: new
Abstract: Bias benchmarks are a popular method for studying the negative impacts of bias in LLMs, yet there has been little empirical investigation of whether these benchmarks are actually indicative of how real world harm may manifest in the real world. In this work, we study the correspondence between such decontextualized "trick tests" and evaluations that are more grounded in Realistic Use and Tangible {Effects (i.e. RUTEd evaluations). We explore this correlation in the context of …

abstract arxiv benchmarks beyond bias cs.cl evaluation harm impacts indicative investigation language language models llms manifest negative popular stat.ap studying tests trick type work world