Why the ability to run AI models on any hardware is becoming a strategically important task and how the open-source community is solving it.
AI: Events
OLMo Hybrid: Transformers and Recurrent Networks Join Forces
Technical context • Research
Allen AI has introduced OLMo Hybrid, an open language model that combines two architectures for more efficient processing of long texts.
Stanford researchers tested leading AI models on their ability to navigate space and found surprisingly poor results.
A new study reveals that cryptocurrency markets defy the fundamental laws of equilibrium, and the culprits aren't errors in risk assessment, but rather more 'human' factors.
AI: Events
25x Inference Speedup: What's Happening with AI Performance on New NVIDIA Hardware
Infrastructure
The new NVIDIA GB300 NVL72 server, paired with the SGLang framework, has demonstrated a 25x performance boost when running language models.
OpenHands has introduced a tool for automatically fixing vulnerabilities in code: an AI agent autonomously analyzes the problem and provides a ready-to-implement solution.
AI: Events
Instant Neural Network Updates: How Doc-to-LoRA and Text-to-LoRA Are Changing the Game
Research
Sakana AI has proposed a method to instantly update the knowledge of language models without costly retraining – by generating adapters directly from text.
Researchers from Allen AI analyzed 250,000 queries to scientific AI tools to uncover how scientists genuinely interact with them in practice.
Perplexity has released two new models for semantic search – designed to quickly and accurately find information across billions of documents.