GRAMA: Adaptive Graph Autoregressive Moving Average Models

M Eliasof, A Gravina, A Ceni, C Gallicchio… - arxiv preprint arxiv …, 2025 - arxiv.org
Graph State Space Models (SSMs) have recently been introduced to enhance Graph Neural
Networks (GNNs) in modeling long-range interactions. Despite their success, existing …

Conic Activation Functions

C Fu, LD Cohen - UniReps: 2nd Edition of the Workshop on Unifying … - openreview.net
Most activation functions operate component-wise, which restricts the equivariance of neural
networks to permutations. We introduce Conic Linear Units (CoLU) and generalize the …

Rethinking Fine-tuning Through Geometric Perspective

KSI Mantri, M Eliasof, CB Schönlieb… - UniReps: 2nd Edition of the … - openreview.net
Fine-tuning pre-trained neural networks has become a cornerstone of transfer learning.
However, the practical success of existing methods like low-rank adaptation (LoRA) lacks …