Intellectual hub of the topic
A new defense system helps browser AI agents recognize malicious instructions hidden on web pages, preventing them from bypassing user tasks.
The Cursor team has granted access to an experimental feature that allows AI to independently handle project code over several iterations without human intervention.
The new DRACO benchmark evaluates how accurately, thoroughly, and objectively AI systems handle complex topic exploration across various fields of knowledge.
Microsoft has introduced a method for detecting hidden vulnerabilities in open-source language models, along with a tool for mass scanning.
AI: Events
Technical context • Infrastructure
AMD explains why true AI autonomy doesn't start with algorithms, but with a sound data strategy and a unified platform to harness it.
A language model designed for working with scientific literature has received recognition from one of the most authoritative journals in the science world.
Chinese company Tencent is releasing a large multimodal model to the public. It is already ranked in the top 7 globally for image editing on LMArena, and it's the best among open models.
LG AI Research shared details about K-EXAONE – a multimodal model developed using proprietary technology, specifically tailored for the Korean language and cultural context.
Anthropic and Apple have reached an agreement: developers can now summon the AI assistant Claude from the code editor – faster and without switching between windows.
Don’t miss a single experiment!
Subscribe to our Telegram channel —
we regularly post announcements of new books, articles, and interviews.