Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Feb 24, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

"What is relevant in a text document?": An interpretable machine learning approach.

Leila Arras¹, Franziska Horn², Grégoire Montavon²

¹Machine Learning Group, Fraunhofer Heinrich Hertz Institute, Berlin, Germany.

|August 12, 2017

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

Prediction of hypertension and restenosis under guideline-directed management in aortic coarctation: development and validation of machine-learning models.

EClinicalMedicine·2026

Same author

Towards robust foundation models for digital pathology.

Nature communications·2026

Same author

Beyond attention heatmaps: How to get better explanations for multiple instance learning models in histopathology.

Medical image analysis·2026

Same author

AI-based discovery of functional boundaries in the human brain from intraoperative electrophysiology.

medRxiv : the preprint server for health sciences·2026

Same author

Modeling attention and binding in the brain through bidirectional recurrent gating.

Nature communications·2026

Same author

Multimodal Deep Learning for Prediction of Progression-Free Survival in Patients with Neuroendocrine Tumors Undergoing <sup>177</sup>Lu-Based Peptide Receptor Radionuclide Therapy.

Cancers·2026

Same journal

Thymidylate synthase inhibitory drugs induce p53-dependent pathways differently.

PloS one·2026

Same journal

Top-down and bottom-up attention for joint pattern classification and reconstruction.

PloS one·2026

Same journal

Short- and long-term scaling behavior of blood pressure and pulse arrival time during sleep in healthy controls and patients with obstructive sleep apnea.

PloS one·2026

Same journal

Double DQN-based secrecy energy efficiency and fairness performance in IRS-assisted NOMA systems with friendly jamming.

PloS one·2026

Same journal

10 recommendations for strengthening citizen science for improved societal and ecological outcomes: A co-produced analysis of challenges and opportunities in the 21st century.

PloS one·2026

Same journal

Paying in public: Peer effects, impression management, and willingness to pay on digital payment platforms.

PloS one·2026

See all related articles

Machine learning models can now categorize text, but understanding their decisions is key. Layer-wise relevance propagation (LRP) explains these predictions by highlighting important words, enhancing model transparency.

Area of Science:

Natural Language Processing
Machine Learning Explainability
Computational Linguistics

Background:

Machine learning (ML) models excel at text categorization, processing vast document collections.
Understanding the reasoning behind ML text classification is crucial for trust and application.
Current methods often lack transparency in how decisions are made.

Purpose of the Study:

To demonstrate how Layer-Wise Relevance Propagation (LRP) can explain ML text classification decisions.
To adapt LRP for word-based ML models like CNNs and SVMs.
To develop new vector-based document representations using word relevance scores.

Main Methods:

Trained convolutional neural network (CNN) and bag-of-words Support Vector Machine (SVM) models for topic categorization.

Related Experiment Videos

Last Updated: Feb 24, 2026

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness

Published on: December 6, 2024

Adapted Layer-Wise Relevance Propagation (LRP) to decompose model predictions onto individual words.

Generated novel document representations based on word-wise relevance scores.

Main Results:

LRP successfully traced classification decisions back to specific words, quantifying their contribution.
Word relevance scores were used to create semantic vector-based document representations.
The CNN model demonstrated higher explainability compared to the SVM, despite similar classification accuracy.

Conclusions:

Layer-Wise Relevance Propagation (LRP) provides a method for understanding and explaining ML-based text categorization.
Novel document representations derived from word relevance enhance semantic understanding.
Explainable AI techniques like LRP are vital for increasing the comprehensibility and utility of ML models in text analysis.