Mendeley Ask My Library (AML)
Last updated on November 25, 2025
For more information on other AI features in Mendeley, detailed FAQs and a technical overview, please visit our Getting started with AI features in Mendeley
The Ask My Library feature provides insights based on your query from across the entirety of the PDF documents that you have stored in your personal Mendeley library.
Get Started
Start your journey by describing what you are looking for in your own words. The system will optimize your query to return the most relevant results, more detailed and in-depth queries will generally result in better responses.
Some suggestions for making your query more in-depth:
- Be specific: Clearly define the scope of your question. Instead of asking a broad question like "What is AI?", you might ask "What are the latest advancements in reinforcement learning algorithms for autonomous vehicles?"
- Contextualize your question: Provide context to help the AI understand the background of your query. For example, "In the context of healthcare, how is machine learning being used to predict patient outcomes?"
- Ask for comparisons: If you're looking for a comparative analysis, frame your question to reflect that. For example, "How do convolutional neural networks compare to recurrent neural networks in image recognition tasks?"
- Seek specific applications: Focus on particular applications or case studies. For example, "What are the applications of natural language processing in legal document analysis?"
- Request detailed explanations: If you need a thorough explanation, make that clear. For example, "Can you explain the key differences between supervised and unsupervised learning with examples?"
Ask My Library processes your query and returns up to 20 relevant documents from your personal library. The Large Language Model (LLM) distills insights from these documents and generates a response, referencing specific passages.
Not all returned documents are used in the summary.
Use the copy functionality to save and store the response.
Technical Details
The Ask My Library AI feature uses a technique called RAG (Retrieval Augmented Generation) to generate reliable responses. It optimizes your query, runs it across a safety check and then uses vector search (semantic similarity) to find relevant text snippets. These snippets are then sent to a Large Language Model (LLM) to generate an answer based on them. The answer will reference the original documents for transparency.
Only the twenty most relevant snippets are sent to the LLM, which decides how many to use for the summary answer.
Limitations
- Each submitted query has a 500-character limit.
- Filtering the content used by collection, or specific documents are not yet supported.
- Non-text content (tables, images, etc.) are not supported.
- Works best on academic papers
- Works on PDF files only
Did we answer your question?
Related answers
Recently viewed answers
Functionality disabled due to your cookie preferences