Discover the Best AI Resources

ai4privacy/pii-masking-300k

Provides 300k annotated multilingual text examples for identifying and masking personally identifiable information (PII) across multiple domains and languages (EN, FR, DE, IT, ES, NL). Intended for training and evaluating token-level PII detection and masking models; includes a DOI for citation.

Cybersecurity AI (CAI)

Framework for building offensive and defensive security agents that run real pentests autonomously. Uses a ReACT loop over 300+ models (OpenAI, Anthropic, DeepSeek, local Ollama) with built-in recon, exploitation, and privilege-escalation tools.

Anthropic's Prompt Engineering Interactive Tutorial

Nine-chapter course teaching prompt engineering for Claude: from basic prompt structure through roles, output formatting, and hallucination control to complete prompts for chatbot, legal, finance, and coding tasks. Runs as editable Jupyter notebooks.

Dive into LLMs (《动手学大模型》系列编程实践教程)

Hands-on coding tutorial series for large language models with slides and runnable notebooks covering fine-tuning, prompting, RLHF, safety, steganography, watermarking, multimodal models, GUI agents, and deployment. Community-maintained, free course materials for students and researchers.

Self-hostable answer engine that pairs SearxNG web search with a local (Ollama) or cloud LLM to return cited answers on your own hardware. An open-source Perplexity alternative with Speed, Balanced, and Quality modes plus image, video, and file search.

tiny-gpu

A minimal GPU written in under 15 SystemVerilog files to teach how GPUs execute parallel kernels from the ground up. Includes an 11-instruction ISA, multiple cores with ALUs and load-store units, a fetch-decode-execute pipeline, and matrix kernels.

Runs a privacy-first, self-hosted answering engine that combines web retrieval with local and cloud LLMs to produce cited answers. Supports SearxNG search, file uploads, image/video search, and mix-and-match models with Speed/Balanced/Quality modes.

An MCP server giving Claude and other AI assistants direct control of the local terminal and file system: run shell commands, manage long-running processes, and search and diff-edit files across the whole OS, not just one project folder.

Firecrawl

Turns any website into clean markdown, structured JSON, or screenshots through a single API — handling JavaScript rendering, rotating proxies, rate limits, and full-site crawling so LLM apps get web data without running scraping infrastructure.

MLX-VLM

Provides local inference, fine-tuning, and a server/CLI for vision–language and omni (image/audio/video) models via MLX. Supports multi-image chat, audio/video inputs, activation quantization (CUDA), TurboQuant KV cache, and LoRA/QLoRA fine-tuning for on-device workflows.

FineWeb

Provides a cleaned, deduplicated English web corpus optimized for LLM pretraining—over 15T tokens aggregated from CommonCrawl with per-dump snapshots and smaller sampled configs (10B/100B/350B). Includes the datatrove processing pipeline, MinHash deduplication, and an ODC-By v1.0 license; suited for large-scale model training and ablation studies but not specialized for code.

Chains pre-trained AI weather and climate models like GraphCast, Pangu, and FourCastNet into composable inference pipelines. Swap prognostic or diagnostic components, plug in reanalysis sources, and add ensemble perturbations or in-loop metrics.