Transfer Learning

SNELL: Sparse tuning with kerNElized LoRA achieves state-of-the-art parameter-efficient fine-tuning performance with drastically reduced memory usage.

Prompt Gradient Alignment (PGA) enhances unsupervised domain adaptation by aligning per-objective gradients in a multi-objective optimization framework, achieving state-of-the-art results.

Provably efficient algorithms for learning with distribution shift are introduced, generalizing and improving prior work by achieving near-optimal error rates and offering universal learners for large…

This study reveals that high input similarity paired with low output similarity is detrimental to continual learning, whereas the opposite scenario is relatively benign; offering insights into mitigat…

Diff-Tuning: a simple yet effective approach transfers pre-trained diffusion models to various downstream tasks by leveraging the ‘chain of forgetting’ phenomenon, improving transferability and conver…

Deep Graph Mating (GRAMA) enables training-free knowledge transfer in GNNs, achieving results comparable to pre-trained models without retraining or labeled data.

ACON: Adversarial CO-learning Networks enhances time series domain adaptation by cleverly combining temporal and frequency features. Frequency features boost within-domain discriminability, while temp…

MolPeg, a novel molecular data pruning framework, enhances model generalization in transfer learning by using a source-free approach and consistently outperforming other methods, even surpassing full-…

Bayesian-guided Label Mapping (BLM) enhances visual reprogramming!

GMDI: a novel Bayesian domain adaptation algorithm significantly improves adaptation by dynamically modeling domain indices using Gaussian Mixture Models, outperforming state-of-the-art methods.

Multi-task learning boosts adversarial robustness in transfer learning by leveraging diverse source data to build a shared representation, enabling effective learning in data-scarce target tasks, as p…