Philip H. S. Torr

This article explores how training data affects model performance in multimodal systems.

An overview of prompting and prefix-tuning in transformer models.

DietCL effectively combines labeled and unlabeled data for better learning.

Exploring the responsible use of generative AI technology in various fields.

Examining how recurrent models can approximate functions based on prompts.

A new method enhances sensor data use in self-driving cars.

A new approach streamlines safety and helpfulness in language model training.

Discover how MALT enhances problem-solving through teamwork among language models.