mteb-retrieve

What problem does it solve?

This skill streamlines the process of finding semantically similar documents or items to a given query using embedding models, crucial for tasks like document ranking and information retrieval.

Core Features & Use Cases

• Data Preprocessing: Cleans raw text data by removing formatting artifacts like line numbers and whitespace.
• Embedding Computation: Guides the selection and use of appropriate embedding models for accurate semantic representation.
• Similarity Ranking: Computes similarity scores and ranks documents to identify top-k relevant results.
• Use Case: When searching a large knowledge base for documents related to "agentic AI patterns," this skill helps identify and rank the most relevant articles based on their semantic meaning.

Quick Start

Use the mteb-retrieve skill to find the top 5 documents semantically similar to the query "best practices for LLM agents" from the provided document set.