Viet-Hoang Tran

Quasi-Equivariant Metanetworks

Tree-sliced Sobolev IPM

Mixed-Curvature Tree-Sliced Wasserstein Distance

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

Discontinuities in Sparse Mixture-of-Experts: A Measure-Stochastic Analysis

Revisiting Tree-Sliced Wasserstein Distance Through the Lens of the Fermat–Weber Problem

Accelerating Transformer Training: Architectural Symmetry, Positional Encoding, and Teleportation

Modeling Expert Interactions in Sparse Mixture of Experts via Graph Structures

Equivariant Metanetworks for Mixture-of-Experts Weights

On Linear Mode Connectivity of Mixture-of-Experts Architectures

Dynamical Properties of Tokens in Self-Attention and Effects of Positional Encoding

Tree-Sliced Entropy Partial Transport