Generate® White Paper
This short white paper explains the basics behind how Generate can help you solve your challenges around producing content and items.
A NLG System for Assessment Content Creation: Generate®
This paper describes Generate in technical detail including the Natural Language Generation (NLG) approach and language modeling architecture that Generate is built on. The graphical user interface (content dashboard) is also covered, as well as results with SMEs where productivity time savings are in the order of 10 fold.
Finetune presented this paper at the14th International Conference on Educational Data Mining (EDM 2021) conference.
An AI-Driven Content Tagging System: Catalog™
This paper covers the technical details of Catalog, our Natural Language Processing (NLP) based content classification system that uses advances in transfer learning approaches to deeply and accurately tag content without using keywords or crosswalks. The modeling architecture is covered as well as results from using Catalog to tag content to the Next Generation Science Standards (NGSS).
Finetune presented this paper at the14th International Conference on Educational Data Mining (EDM 2021) conference.
Injecting Knowledge Graph Triples into Large Language Models for Passage & Question Generation
This white paper from the Finetune R&D team details one proposed solution for language model confabulation, otherwise known as “hallucinations”. Rather than relying on the model to recall relevant facts from its training data, the method instead grounds the model on a human-legible graph of semantic relationships.