Hey, in this issue: Services: The new software; Google DeepMind partners with EVE Online for AI model testing; 5 new ways to explore the web with generative AI in Search; What happens when your agent reads before it codes; Multimodal embedding & reranker models with Sentence Transformers; Why SSMs struggle in parameter golf: A structural analysis at 25M parameters; "What I learned building attention residuals from scratch"; and more.
[Sequoia] Services: The new software
[Ars Technica] Google DeepMind partners with EVE Online for AI model testing
[Google] 5 new ways to explore the web with generative AI in Search
Research-driven agents: What happens when your agent reads before it codes
[Tutorial] Multimodal embedding & reranker models with Sentence Transformers
[ChapterPal] Efficient Exploration at Scale from Google DeepMind
Why SSMs struggle in parameter golf: A structural analysis at 25M parameters
[Ai2] Train separately, merge together: Modular post-training with mixture-of-experts
Support my regular work on curating this ad-free newsletter by becoming a Cool Supporter 🚀
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.