Profiles2020-2025
Zhi-Quan Luo
Shenzhen Research Institute of Big Data
10Papers
7Last author
26Unique co-authors
0First author
0Solo
First and last author come from byline position, which some venues assign alphabetically.
Recent papers10
- Preserving Diversity in Supervised Fine-Tuning of Large Language ModelsAuthor 6 of 7ICLR 2025
- Why Transformers Need Adam: A Hessian PerspectiveAuthor 6 of 6NeurIPS 2024 · 149 citations
- Provable Adaptivity of Adam under Non-uniform SmoothnessAuthor 4 of 5KDD 2024 · 4 citations
- Towards Memory- and Time-Efficient Backpropagation for Training Spiking Neural NetworksAuthor 6 of 6ICCV 2023 · 84 citations
- Imitation Learning from Imperfection: Theoretical Justifications and AlgorithmsAuthor 5 of 6NeurIPS 2023 · 26 citations
- PAC-Bayesian Spectrally-Normalized Bounds for Adversarially Robust GeneralizationAuthor 3 of 3NeurIPS 2023 · 14 citations
- Adam Can Converge Without Any Modification On Update RulesAuthor 5 of 5NeurIPS 2022 · 134 citations
- Stability Analysis and Generalization Bounds of Adversarial TrainingAuthor 5 of 5NeurIPS 2022 · 49 citations
- When Expressivity Meets Trainability: Fewer than Neurons Can WorkAuthor 5 of 5NeurIPS 2021 · 11 citations
- A Single-Loop Smoothed Gradient Descent-Ascent Algorithm for Nonconvex-Concave Min-Max ProblemsAuthor 4 of 4NeurIPS 2020 · 130 citations
