engblogs

This blogpost discusses the challenges of reasoning on charts in visual language and proposes a foundation model called MatCha, which is trained on chart de-rendering and math reasoning tasks, surpassing previous state-of-the-art models in ChartQA.

How to incorporate democratic inputs into AI development

Exploring the history, capabilities, and limitations of large language models, from the emergence of ChatGPT to the evolution of models like BERT and GPT-3, discussing key terms like attention, embeddings, and transformers, along with ways to adapt and use these models for specific tasks.

A research framework for evaluating general-purpose AI models against novel threats, including extreme risks, to ensure responsible development and deployment.

Probabilistic AI that provides accurate estimations of data explanations.

Using AI to discover a potential drug to combat drug-resistant infections by identifying a compound that kills Acinetobacter baumannii.

Upgrade to Reimagine XL by Clipdrop for an ultimate image reimagination tool with high-quality output, various aspect ratio handling, and no technical know-how needed.

The blogpost discusses the concept of multiple layers of abstraction in design systems and their advantages and disadvantages.

The blog post discusses practical end-to-end diversification in search and recommender systems, focusing on representation online and the importance of diversity. It explores how the diversification process works, the components involved, and techniques such as Round Robin and Determinantal Point Process for diversifying rankings. It also covers techniques for diversification at the retrieval stage, including Overfetch-and-Rerank and Bucketized ANN retrieval.

A program to fund experiments in establishing a democratic process for determining rules for AI systems within legal limits.

The blogpost highlights the research efforts behind some of the announcements made at Google I/O 2023, including advancements in language models like PaLM 2, image generation and editing models like Imagen, and text-to-video generation models like Phenaki.

This blogpost discusses differentially private clustering algorithms for large-scale datasets, focusing on hierarchical graph clustering and metric space clustering. It also explores the application of differentially private clustering in a real-world scenario for publishing COVID vaccine-related queries while protecting user privacy.