Fuyang (Scott) Cui
Computer Science @ University of Toronto
Chronological list of papers, preprints, and workshop manuscripts. Bold indicates my authorship. * indicates equal contribution.
Kunlun Zhu, Zijia Liu, Bingxuan Li, Muxin Tian, Yingxuan Yang, Jiaxun Zhang, Pengrui Han, Qipeng Xie, Fuyang Cui, Weijia Zhang, Xiaoteng Ma, Xiaodong Yu, Gowtham Ramesh, Jialian Wu, Zicheng Liu, Pan Lu, James Zou, Jiaxuan You
Taxonomy, benchmark, and debugging framework that isolates cascading failures in LLM agents and uses targeted feedback to improve reliability.
Chandler Smith, Marwa Abdulhai, Manfred Diaz, Marko Tesic, Rakshit Trivedi, Fuyang Cui, et al.
Evaluates how well LLM agents generalize in zero-shot mixed-motive social scenarios using the Concordia simulation benchmark.
Blair Yang *, Fuyang Cui *, Keiran Paster, Jimmy Ba, Pashootan Vaezipoor, Silviu Pitis, Michael R. Zhang
An automated qualitative-evaluation framework for specialized, open-ended, and agentic tasks of LLMs.
Yuhan Chen *, Fuyang Cui *, Shouyi Hung *, Liwei Yang *
Augments DreamerV1 with an auxiliary imitation loss to improve world-model imagination accuracy.