Machine Learning & Deep LearningFlashAttention is Tri Dao's GPU-memory-aware attention algorithm delivering 2-4x speedup and long-context training. FA-3 reaches 75% of H100 FLOPS. Architecture, patterns, comparison.
Machine Learning & Deep LearningvLLM is an open-source LLM serving engine using PagedAttention for near-100% GPU memory utilization and up to 24x throughput. Architecture, patterns, and TGI/TensorRT-LLM comparison.
Generative AI & LLMsSora 2 is OpenAI's second-generation video AI with 1080p output, synced audio, and 25-second clips. Architecture, pricing, and how it differs from Sora 1 and Veo 3.
Machine Learning & Deep LearningThe KV Cache is the optimization that turns Transformer LLM inference from quadratic to linear by reusing computed Key and Value tensors. Learn how it works, its memory cost, and how it relates to Prompt Caching, PagedAttention, and vLLM.
Generative AI & LLMsManus is an autonomous AI agent built by Butterfly Effect (Singapore, founded in China). Learn its multi-agent architecture, how it differs from Claude and ChatGPT, and the April 2026 events that blocked Meta's $2B acquisition.
Generative AI & LLMsThe Bash Tool is Anthropic's built-in API tool that lets Claude execute shell commands in a persistent bash session. Learn how it works, how it differs from Computer Use, and the canonical implementation pattern with Python examples.
Generative AI & LLMsIn-context Learning (ICL) is when an LLM picks up a new task from examples shown in the prompt itself. This guide covers zero-shot, few-shot, chain-of-thought, fine-tuning comparisons, and production patterns.
Machine Learning & Deep LearningQuantization compresses LLM weights into lower-precision formats to save memory and speed up inference. This guide covers FP16/INT8/INT4, GPTQ, AWQ, GGUF, hallucination trade-offs, and production patterns.
Generative AI & LLMsHallucination in LLMs is when a model produces plausible but incorrect output. This guide covers the causes, taxonomy, detection techniques, RAG and Constitutional AI mitigations, and the latest 2026 research.
Generative AI & LLMsA multi-agent system is an architecture where several LLMs collaborate to solve a task. Learn orchestrator-worker patterns, real-world examples, and Anthropic's research configuration.
