Evaluating LLMs' ability to detect harmful instructions in user prompts.
― 6 min read
Xifeng Yan
A new approach enhances task-oriented dialogue systems using function calling.
― 6 min read
High-quality image-text pairs improve performance of multimodal models in various tasks.
― 5 min read
A dataset to improve AI's ability to read advanced scientific materials.
― 6 min read