Controlling for Stereotypes in Multimodal Language Model Evaluation
원문 초록
We propose a methodology and design two benchmark sets for measuring to what extent language-and-vision language models use the visual signal in the presence or absence of stereotypes. The first benchmark is designed to test for stereotypical colors of common objects, while the second benchmark considers gender stereotypes. The key idea is to compare predictions when the image conforms to the stereotype to predictions when it does not. Our results show that there is significant variation among multimodal models: the recent Transformer-based FLAVA seems to be more sensitive to the choice of image and less affected by stereotypes than older CNN-based models such as VisualBERT and LXMERT. This effect is more discernible in this type of controlled setting than in traditional evaluations where we do not know whether the model relied on the stereotype or the visual signal.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
Enterprise Large Language Model Evaluation Benchmark
Finance Language Model Evaluation (FLaME)
Making Language Model a Hierarchical Classifier and Generator
CoLMbo: Speaker Language Model for Descriptive Profiling
Data Efficacy for Language Model Training
Research on Model Parallelism and Data Parallelism Optimization Methods in Large Language Model-Based Recommendation Systems
토론
이 논문과 연결된 포럼 글이 아직 없습니다.