← Search

Giacomo Salici

2 accepted papers

2026

Gradient-Sign Masking for Task Vector Transport Across Pre-Trained Models

ICLR 2026poster

When a new release of a foundation model is published, practitioners typically need to repeat fine-tuning, even if the same task was already tackled in the previous version. A promising alternative is to reuse the parameter changes (i.e., task vectors) that capture how a model adapts to a specific t…

Cited by 0SourcecodeScholar
2026

Transporting Task Vectors across Different Architectures without Training

ICML 2026poster

Adapting large pre-trained models to downstream tasks often produces task-specific parameter updates that are expensive to relearn for every model variant. While recent work has shown that such updates can be transferred between models with identical architectures, transferring them across models of…

Cited by 0SourceScholar