A new watermarking method safeguards customized language models from unauthorized use.
― 5 min read
Yaliang Li
Discover how ChartThinker enhances chart summaries for better understanding.
― 6 min read
A new method for combining training data enhances language model performance.
― 5 min read
FedBiOT improves large language models while keeping data private and resource usage low.
― 5 min read
New methods generate synthetic data to protect privacy in collaborative scenarios.
― 6 min read
This method improves safety in image generation while maintaining quality.
― 6 min read
A look at improving image creation from text descriptions.
― 5 min read