Archive - The Update of The Gradient

New

Top

Discussion

Mini-Update #30: OpenAI Updates and GAIA

Sam Altman returns to OpenAI, and researchers develop a benchmark for general AI assistants.

Nov 28 •

and

Update #63: AI Brings Oxygen to Mars and Pretraining Data Mixtures

A break from the drama

Nov 21 •

,

, and

Mini-Update #29: Grok and NOIR

Elon Musk's xAI releases its first AI model, and Stanford researchers propose a system that allows humans to control robots using brain signals.

Nov 14 •

and

Update #62: The Generative AI Paradox and New AI Bills

Biden releases an Executive Order on AI; Rishi Sunak unveils a landmark deal on AI testing; researchers show that AI models' generative capabilities do…

Nov 7 •

and

October 2023

Update #61: AI in Smartphone Cameras and MatFormer

We consider the implications of automated photo editing; researchers improve transformer inference efficiency.

Oct 24 •

,

, and

Mini-Update #28: Problems with AI Chatbots and Language Models as Analogical Reasoners

Meta's and Character AI's chatbot launches are rife with setbacks; researchers propose a new method to guide LLMs' reasoning processes.

Oct 17 •

and

Update #60: Wearable AI and Efficient Streaming Language Models

A slew of competitors announce AI-powered wearables, and researchers improve the performance of streaming language models by exploiting attention sinks.

Oct 10 •

,

, and

Mini-Update #27: Meta's ChatGPT Competitor and ViTs' Need for Registers

Meta rolls out new AI features across its platforms, and researchers improve Vision Transformer performance with register tokens.

Oct 3 •

and

September 2023

Update #59: Generative AI in the Classroom and Next-Token Predictors as Universal Learners

We consider the benefits and drawbacks of using generative AI to help teachers and students; Eran Malach argues that auto-regressive next-token…

Sep 26 •

,

, and

Mini-Update #26: Schumer's AI insight forum and PagedAttention

Senator Chuck Schumer hosts the first bipartisan federal AI forum, and researchers develop an optimization to make LLMs more efficient and effective.

Sep 19 •

and

Update #58: Hidden Meme Messages & LLMs as Optimizers

New work develops AI techniques for encoding hidden messages in innocuous media, and a DeepMind team develops a framework for using LLMs to solve…

Sep 12 •

,

,

, and

Mini-Update #25: ChatGPT Enterprise and StableVideo

OpenAI launches a ChatGPT service for corporate use, and researchers at Microsoft release a text-driven framework for video editing.

Sep 5 •

and

#nojs-banner { position: fixed; bottom: 0; left: 0; padding: 16px 16px 16px 32px; width: 100%; box-sizing: border-box; background: red; color: white; font-family: -apple-system, "Segoe UI", Roboto, Helvetica, Arial, sans-serif, "Apple Color Emoji", "Segoe UI Emoji", "Segoe UI Symbol"; font-size: 13px; line-height: 13px; } #nojs-banner a { color: inherit; text-decoration: underline; } This site requires JavaScript to run correctly. Please turn on JavaScript or unblock scripts