A.I. This AI research dives into the limitations and capabilities of large language models (LLMs) of transformers, empirically and theoretically, in composition tasks. 06/04/2023
A.I. Meet StyleAvatar3D: A new AI method for generating stylized 3D avatars using image and text diffusion models and a GAN-based 3D rendering network 06/04/2023
A.I. Researchers at Imperial College London propose FitMe: an AI model that turns arbitrary facial images into relightable facial avatars, which can be used directly in common rendering engines and games 06/04/2023
A.I. UT Austin and UC Berkeley researchers present Ambient Diffusion: an AI framework for training/tuning diffusion models given only corrupted data as input 06/04/2023
A.I. How should we maximize the planning capacity of LLMs while reducing the cost of computing? Meet SwiftSage: a novel generative agent for complex interactive reasoning tasks, inspired by the dual process theory of human cognition 06/03/2023
A.I. Do you really need reinforcement learning (RL) in RLHF? New Stanford research proposes DPO (direct preference optimization): a simple training paradigm for training language models from preferences without RL 06/03/2023
A.I. Researchers from Stanford, UC Berkeley, and Adobe Research have developed a new AI model that can realistically insert specific humans into different scenes. 06/03/2023
A.I. Unleashing the Power of Visual Language: Google AI Proposes MatCha and DePlot to Revolutionize Graph Comprehension and Mathematical Reasoning 06/03/2023
