Tag
17 articles
This article explains time series foundation models, focusing on Mistral AI's Timestral model that outperforms Google's TimesFM-3 while running efficiently on CPUs.
This article explains the technical advancements behind Anthropic's Claude Sonnet 5.5, focusing on how improved efficiency and token usage reduce costs while maintaining high performance.
Anthropic has released Sonnet 5.5, a more affordable and efficient mid-range AI model with faster response times and reduced token consumption.
OpenRouter's token consumption has surged 25,000% since early 2025, sparking debate over the sustainability of current AI development practices and raising concerns about a potential AI bubble.
Learn how model distillation works - a technique that allows large, complex AI models to teach smaller, more efficient models how to perform the same tasks. This makes AI more accessible and practical for everyone.
OpenAI's GPT-5.6 enhances AI efficiency across models, inference, and agentic workflows, delivering more useful intelligence per dollar. The update marks a significant step toward sustainable and cost-effective AI development.
Developers have successfully deployed the 1-bit Bonsai-27B language model using PrismML’s fork of llama.cpp, enabling efficient local inference with OpenAI-compatible workflows.
This article explains Mixture of Experts (MoE) AI models, how they work like teams of specialists, and why they're important for efficient AI performance.
Learn how Baidu's Unlimited OCR achieves efficient processing of dozens of document pages in a single pass by mimicking human memory and forgetting mechanisms.
This explainer explores how Un0, a new AI system from Databricks' former AI chief, achieves 1,000x energy efficiency improvements through novel architectural approaches that could revolutionize AI development and deployment.
Google DeepMind introduces DiffusionGemma, a 26 billion parameter open model using text diffusion to achieve up to 4x faster text generation on GPUs.
Learn how a new AI system enables faster and more efficient continual learning by running multiple experiments at once using LoRA adapters.