What Is Dense Passage Retrieval?
Dense passage retrieval encodes questions and passages as vectors and retrieves passages whose learned representations are close to the query.
Don't Just Read it – Explore It
Dense passage retrieval encodes questions and passages as vectors and retrieves passages whose learned representations are close to the query.
Document chunking divides source material into retrievable units whose boundaries and context strongly affect what a RAG system can find.
Long-context retrieval identifies and emphasizes the most useful information within documents or prompts that are too long to treat every part equally.
ColBERT represents queries and documents as sets of token-level vectors and scores them with efficient late interactions.
Context compression reduces the amount of text supplied to a model while trying to preserve the information most important for the task.
Retrieval routing directs a query to the search source or retrieval method most likely to contain useful, permitted evidence.
A multi-agent system uses several autonomous or semi-autonomous agents that communicate, specialize, cooperate, or compete within a shared environment.
Agentic RAG lets an AI system decide when, where, and how to retrieve information and whether the evidence is sufficient before answering.
An AI orchestrator coordinates models, tools, data, and workflow state so each part of a larger automated process runs at the right time.
Task decomposition breaks a complex objective into smaller subproblems that can be solved, checked, and combined more systematically.
An AI agent loop repeatedly observes the current state, chooses an action, receives a result, and updates what it should do next.
Agent planning is the process of turning a goal into an ordered or branching set of actions that an AI system can revise as conditions change.