10 papers · 1 filter
Compositional Generalization in Autoregressive Models via Logit Composition
Aakash Kumar, Maria Sofia Bucarelli, Emanuele Natale
Composing autoregressive models remains a core challenge in understanding how large language models can combine behaviors or skills learned across tasks. We introduce a new and pri…
MASS: MoErging through Adaptive Subspace Selection
Donato Crisostomi, Alessandro Zirilli, Antonio Andrea Gargiulo +5
Model merging has recently emerged as a lightweight alternative to ensembling, combining multiple fine-tuned models into a single set of parameters with no additional training over…
Energy Guided smoothness to improve Robustness in Graph Classification
Farooq Ahmad Wani, Maria Sofia Bucarelli, Andrea Giuseppe Di Francesco +2
Graph Neural Networks (GNNs) are powerful at solving graph classification tasks, yet applied problems often contain noisy labels. In this work, we study GNN robustness to label noi…
Early-Exit Graph Neural Networks
Andrea Giuseppe Di Francesco, Maria Sofia Bucarelli, Franco Maria Nardini +3
Early-exit mechanisms allow deep neural networks to stop inference once prediction confidence is high, reducing latency and energy on easy inputs while retaining full-depth accurac…
PISA: Prioritized Invariant Subgraph Aggregation
Ali Ghasemi, Farooq Ahmad Wani, Maria Sofia Bucarelli +1
Recent work has extended the invariance principle for out-of-distribution (OOD) generalization from Euclidean to graph data, where challenges arise due to complex structures and di…
Subtract the Corruption: Training-Data-Free Corrective Machine Unlearning using Task Arithmetic
Mostafa Mozafari, Farooq Ahmad Wani, Maria Sofia Bucarelli +1
Corrupted training data are ubiquitous. Corrective Machine Unlearning (CMU) seeks to remove the influence of such corruption post-training. Prior CMU typically assumes access to id…