Tag
36 articles
Sakana AI has launched Sakana Fugu, an orchestration model that dynamically routes tasks across a swappable pool of frontier LLMs, achieving strong performance on coding, reasoning, and agentic benchmarks.
Learn to build a simple multi-LLM orchestration system that dynamically selects the best AI model for each question, similar to Sakana AI's Fugu system.
This explainer introduces the Open Knowledge Format (OKF), a vendor-neutral specification for structuring knowledge for AI agents. It explains how OKF uses Markdown and YAML to create curated knowledge bundles, distinguishing it from traditional RAG systems.
NVIDIA's garak framework offers a comprehensive solution for defensive LLM red-teaming, enabling organizations to identify vulnerabilities and enhance model safety.
This article explains OpenAI's new Dreaming memory system that builds coherent user profiles from conversations, improving information retention from 52.2% to 75.1%.
This article explains how no-code AI agent builders like Stack AI work and why they're transforming enterprise workflow automation by enabling non-technical users to create sophisticated autonomous AI systems.
Researchers from NUS, MIT, and A*STAR introduce MEMO, a modular framework that enables LLMs to learn new knowledge without modifying core parameters.
EAGLE 3.1, developed by the EAGLE team, vLLM, and TorchSpec, tackles attention drift in LLM inference, enhancing speculative decoding stability for production use.
Together AI open-sources OSCAR, an attention-aware 2-bit KV cache quantization system that significantly reduces memory usage and improves decoding speed for long-context LLMs.
This article explains how to build a complete Langfuse observability and evaluation pipeline for LLM development, covering tracing, prompt management, scoring, and experimentation.
Poetiq's new meta-system automatically builds a model-agnostic inference harness that improves performance across multiple LLMs without fine-tuning.
A new tutorial explores how to build a cost-aware LLM routing system using NadirClaw, which classifies prompts locally and switches between models like Gemini for optimal performance and cost efficiency.
