Recently Added
3,528 documents
四、随机化 SVD
Teaches randomized SVD through three concrete applications: word vector projection, video background removal, and topic modeling.
PRD — NoiseCancel v2: Multi-Platform Feed Intelligence
Plans a multi-platform social feed scraper with semantic dedup, AI digest, and Flutter archive view.
Infinite-Context Chat Storage Specification
Specifies a graph-database-backed chat storage system that uses vector embeddings and adaptive indexing to give LLMs unlimited context without token waste.
Technical Book Rack RAG — Implementation Plan
Plans a 10-phase fork of a PDF RAG system into a technical-book RAG with concept-aware chunking and book metadata.
Makemore Notes
Documents a character-level language model walkthrough covering bigram, MLP, backprop, and Wavenet with practical training observations.
WARP.md
Documents the architecture, data model, and development workflow for an AI memory service with MCP bridge integration.
RAG para Presupuestos de Obra - Plan Tecnico v2.0
Defines a RAG system for Spanish construction budgets, ingesting PDF, TXT, CSV, DOCX, XLSX, and BC3 files to answer price and material queries with fallback market estimates.
2Vect in Haskell - Linear algebra of types, 2-Vect,
Explores type-level matrices, monoidal categories, and the Kronecker product to model transition systems and vector spaces in Haskell.
AIAssist API Documentation
Documents a Python library that unifies AI model interactions, data analysis, image processing, and Git repository management via MCP.
Vitamix POC -- Technical Specification
Defines the architecture, services, AI models, and deployment pipeline for an AI content generation platform on AEM Edge Delivery and Google Cloud.
Cross-ISA Semantic Code Embedding: A Complete Implementation Blueprint
Describes how to build a cross-ISA binary embedding system using IR lifting and contrastive learning, with architecture, training, and deployment details.
Brainy - Smart Bookmark Vault Specification
Specifies a single-user bookmark knowledge base with AI ingestion, hybrid search, knowledge graph, and natural language Q&A.
embedx Specification
Defines an embedding service with an Ollama-compatible HTTP API using Go and Python FastEmbed subprocess with pipe communication.
AI Code Agent - Project Documentation
Documents a complete RAG-powered semantic code search and question-answering backend built with FastAPI, PostgreSQL, and OpenAI.
Word2Vec Embeddings Quiz
Tests understanding of Word2Vec concepts, gensim training, and a product recommender system built on transaction data.
Graph Matching with Topological Features
Teaches enhanced graph matching by combining spatial distances with node2vec and commute times embeddings, then applying the Hungarian algorithm.
WARP.md
Documents a RAG service architecture, development commands, and component extension patterns for a medical device troubleshooting tool.
Solution Approach: Intelligent SHL Assessment Recommendation System
Describes a RAG-based system that recommends SHL assessments from natural language queries using FAISS and LLM re-ranking.
draft
Describes a permutation-invariant neural network architecture that processes sensory inputs in any order using an attention mechanism.
Core principles
Defines 7 Jina AI APIs with endpoints, schemas, examples, and integration rules for building search and classification pipelines.
Memory Connector - Application Specification
Specifies a personal memory management system with AI enrichment, offline sync, and cost controls.
Video Search Backend for CMU Database Course
Builds a semantic video search backend that extracts transcripts, generates local embeddings, and stores them in LanceDB for query-time frame extraction.
QWEN.md - MediaWiki Graph Analysis Platform
Documents architecture, setup, configuration, and conventions for a MediaWiki graph analysis toolkit using embeddings and clustering.
RAG (Retrieval Augmented Generation) Implementation - Technical Outline
Outlines RAG pipeline components, query translation techniques, indexing strategies, and adaptive systems for combining custom data with LLMs.