Ngoc Tan Lai — accepted AI-conference papers

2026

Conservation Laws for Modern Neural Architectures

Viet Hoang Tran, VINH KHANH BUI, Ngoc Tan Lai, Nam Nguyen, Tuan Dam, Tan Nguyen

ICML 2026spotlight

Understanding gradient descent dynamics is key to explaining the success of over-parameterized models, where implicit bias manifests through conservation laws in gradient flow. While such laws are well understood for linear and ReLU networks, they remain largely unexplored for modern architectures. …

Cited by 0 Source ↗Scholar ↗

2026

Functional Equivalence in Attention: A Comprehensive Study with Applications to Linear Mode Connectivity

Viet Hoang Tran, VINH KHANH BUI, Van-Hoan Trinh, Ngoc Tan Lai, Tan Nguyen

ICML 2026poster

Neural network parameter spaces are inherently non-injective, as distinct parameter configurations can realize identical functions through functional equivalence. While this symmetry is well understood in classical fully connected and convolutional models, it becomes substantially more intricate in …

Cited by 0 Source ↗Scholar ↗