Concepts are the durable ideas the AI news keeps mapping onto — the things still worth understanding after the week’s releases are forgotten. GROUNDING links every scored signal to the concepts it touches, so this hub is where the stream turns into a learning map. They cluster into four groups.
This is the index concept page (definitions and fundamentals). For dated developments, see the index Topic hub.
Prompt Engineering is the practice of shaping instructions, examples, constraints, and output formats for LLM behavior. GROUNDING tracks prompt patterns as part of broader context engineering and agent design.
Agent Memory is the set of mechanisms that lets AI agents persist useful context across turns, sessions, tools, and tasks. GROUNDING tracks memory stores, compression, retrieval boundaries, and failure modes for long-running agent workflows.
AI agents are LLM-driven systems that can plan, call tools, inspect results, and iterate toward a goal. GROUNDING tracks agent architectures, benchmarks, tool use, browser agents, coding agents, and reliability patterns.
Chunking is the practice of splitting documents, code, or data into retrieval units for RAG and indexing. GROUNDING tracks chunk size, boundaries, metadata, context windows, and when chunking harms answer quality.
Code agents are AI agents that inspect repositories, edit files, run tools, and iterate on software tasks. GROUNDING tracks coding-agent workflows, context management, evaluations, and reliability risks.
Codebase indexing is the process of making repository structure, symbols, files, and dependencies searchable for AI coding tools. GROUNDING tracks indexing methods, retrieval quality, and agent context design.
Context Engineering is the discipline of selecting, compressing, ordering, and refreshing the information an AI system sees at inference time. GROUNDING tracks context windows, memory, tools, retrieval, and prompt architecture.
Embeddings are vector representations of text, code, images, or other data that make semantic search and clustering possible. GROUNDING tracks embedding models, retrieval quality, evaluation, and production RAG tradeoffs.
Hybrid Search combines lexical search with vector retrieval so RAG systems can match both exact terms and semantic meaning. GROUNDING tracks ranking, recall, query rewriting, and retrieval robustness.
LLM Evals are tests and measurement systems for model behavior, agent performance, retrieval quality, safety, and task success. GROUNDING tracks eval benchmarks, failure modes, and practical measurement patterns.
Long Context refers to model and system designs that process large input windows while preserving useful attention and answer quality. GROUNDING tracks context limits, compression, memory, and retrieval-vs-context tradeoffs.
MCP, the Model Context Protocol, is a standard interface for connecting AI assistants to tools, data, and services. GROUNDING tracks MCP servers, tool integrations, security concerns, and agent workflow patterns.
Open Source LLMs are language models released with weights, code, or permissive access that builders can inspect, fine-tune, or self-host. GROUNDING tracks open model releases, inference stacks, and benchmarks.
RAG, or retrieval-augmented generation, grounds LLM answers in retrieved documents, code, or data before generation. GROUNDING tracks RAG architecture, retrieval quality, evaluation, grounding failures, and production patterns.
Reranking reorders retrieved candidates after an initial search step so the best evidence reaches the LLM first. GROUNDING tracks cross-encoders, LLM rerankers, latency tradeoffs, and RAG quality gains.
Tool Use lets AI systems call external functions, APIs, browsers, shells, or databases instead of only generating text. GROUNDING tracks tool schemas, agent loops, safety boundaries, and reliability patterns.
A Vector Database stores and searches embedding vectors for semantic retrieval, recommendations, and RAG. GROUNDING tracks vector indexes, filtering, hybrid retrieval, latency, and production reliability.
