Hi. I'm Yiwei Yang.
I am a PhD student at the University of Washington, advised by Bill Howe. I received my undergraduate degree in Computer Science in 2019 from the University of Michigan.
I am interested in making our models more reliable and trustworthy. Currently, I am working on building robust reward models to tackle reward hacking of large language models. Recently, I worked on benchmarking and mitigating spurious correlations of Large Multi-modal Models (LMMs).
Selected Publications
- Y. Yang, C. P. Lee, S. Feng, D. Zhao, B. Wen, A. Z. Liu, Y. Tsvetkov, B. Howe. Escaping the SpuriVerse: Can Large Vision-Language Models Generalize Beyond Seen Spurious Correlations? ICML 2025 R2-FM, In Submission to NeurIPS 2025
- Y. Yang, A. Liu, R. Wolfe, A. Caliskan, B. Howe. Label-Efficient Group Robustness via Out-of-Distribution Concept Curation CVPR 2024
- B. Han, Y. Yang, A. Caspi, B. Howe. Towards Zero-shot Annotation of the Built Environment with Vision-Language Models SIGSPATIAL 2024
- R. Wolfe, S. Issac, B. Han, B. Wen, Y. Yang, L. Rosenblatt, B. Herman, E. Brown, Z. Qu, N. Weber, B. Howe. Laboratory-scale AI: Open-Weight Models are Competitive with ChatGPT Even in Low-Resource Settings FAccT 2024
- Y. Yang, B. Howe. Does a Fair Model Produce Fair Explanations? Relating Distributive and Procedural Fairness. HICSS 2024
- Y. Yang, A. Liu, R. Wolfe, A. Caliskan, B. Howe. Regularizing Model Gradients with Concepts to Improve Robustness to Spurious Correlations ICML SCIS 2023
- R. Wolfe, Y. Yang, B. Howe, A. Caliskan. Contrastive Language-Vision AI Models Pretrained on Web-Scraped Multimodal Data Exhibit Sexual Objectification Bias FAccT 2023