AI research

DiffusionAvatars generates realistic 3D avatars

Summary Researchers at the Technical University of Munich have developed DiffusionAvatars, a method for creating high-quality 3D avatars with realistic facial expressions. The system was trained using RGB videos and 3D meshes of human heads. After training, the system is able to animate avatars both by taking animations from the input videos and by generating …

DiffusionAvatars generates realistic 3D avatars Read More »

Multimodal models are easy to confuse, say researchers

Summary A new study by Chinese researchers shows how easy it is to bypass the safety mechanisms of multimodal AI models (MLLM). The study tested the safety of Google Bard and GPT-4V using targeted attacks. Specifically, images were manipulated to deliberately mislead the models (image embedding attack) and to respond to requests that should have …

Multimodal models are easy to confuse, say researchers Read More »

GPT-4 shines in Microsoft radiology study, outperforming human experts on some tasks

Summary Microsoft recently published a study that explores the capabilities and limitations of GPT-4 in radiology. Working with a radiologist and Nuance, a Microsoft company whose PowerScribe solution is used by more than 80 percent of radiologists in the U.S., the research team created a comprehensive evaluation and defect analysis framework. Within this framework, the …

GPT-4 shines in Microsoft radiology study, outperforming human experts on some tasks Read More »

Meta’s AI lab turns 10 with three new AI projects and an impressive demo

Summary To mark the 10th anniversary of Meta’s Fundamental AI Research (FAIR) team, the company presents three new research projects: Ego-Exo4D, Seamless Communication, and Audiobox. Ego-Exo4D is a dataset and benchmark set to support AI research in video learning and multimodal perception. Collected over two years by Metas FAIR, Project Aria, and 15 university partners …

Meta’s AI lab turns 10 with three new AI projects and an impressive demo Read More »

OpenAI CEO Sam Altman comments on the no longer secret Q* project

Summary OpenAI’s “Q*” project was quickly labeled a secret AGI project. Now, returning OpenAI CEO Sam Altman weighs in. Altman indirectly confirms Q without giving any details about the project. When asked by The Verge’s Alex Heath what Q* was about, Altman replied that it was an “unfortunate leak” that he did not want to …

OpenAI CEO Sam Altman comments on the no longer secret Q* project Read More »

Deepmind’s GNoME AI tool speeds up crystal research by 800 years

Summary Discovering new crystal structures is a tedious task for scientists. A new AI tool from Google Deepmind aims to accelerate the process. Google Deepmind has published an article in Nature about its AI tool GNoME, which has discovered more than 2.2 million new crystals. According to Deepmind, these include some 380,000 particularly stable compounds …

Deepmind’s GNoME AI tool speeds up crystal research by 800 years Read More »

Deepmind’s new prompting method takes a step back for more accuracy

Summary A recent paper from Alphabet’s AI company Google Deepmind shows that a simple tweak to prompts can significantly improve the accuracy of large language models. The technique taps into the human ability to abstract. Step-back prompting asks the LLM a general question before the actual task. This allows the system to retrieve relevant background …

Deepmind’s new prompting method takes a step back for more accuracy Read More »

GPT-4 fails at simple tasks that humans can easily solve

Summary Researchers from Metas AI Research (FAIR), HuggingFace, AutoGPT, and GenAI present the GAIA (General AI Assistants) AI benchmark, which measures AI performance on tasks that are easy for humans to solve. The benchmark is based on the hypothesis that a potential General Artificial Intelligence (AGI) must outperform humans even on tasks that are easy …

GPT-4 fails at simple tasks that humans can easily solve Read More »

MechGPT opens up new research opportunities by connecting the dots between unlinked knowledge

Summary Recognizing correlations is one of the core capabilities of an AI model. Specialized language models use this to show connections between different areas of research. Markus J. Buehler, a researcher at the Massachusetts Institute of Technology (MIT), presents a strategy called “MechGPT” that was developed specifically for the study of material failure. But MechGPT …

MechGPT opens up new research opportunities by connecting the dots between unlinked knowledge Read More »

Scroll to Top