Mengya Gao — accepted AI-conference papers

2026

Language-based Trial and Error Falls Behind in the Era of Experience

Haoyu Wang, Guozheng Ma, Shugang Cui, Yilun Kong, Haotian Luo, Li Shen +4

ICML 2026poster

While Large Language Models (LLMs) excel in language-based agentic tasks, their applicability to unseen, nonlinguistic environments (e.g., symbolic or spatial tasks) remains limited. Previous work attributes this performance gap to the mismatch between the pretraining distribution and the testing di…

Cited by 0 Source ↗Scholar ↗

2019

P2SGrad: Refined Gradients for Optimizing Deep Face Models

Xiao Zhang, Rui Zhao, Junjie Yan, Mengya Gao, Yu Qiao, Xiaogang Wang +1

CVPR 2019poster

Cosine-based softmax losses significantly improve the performance of deep face recognition networks. However, these losses always include sensitive hyper-parameters which can make training process unstable, and it is very tricky to set suitable hyper parameters for a specific dataset. This paper add…

Cited by 48 PDF ↗Scholar ↗