This submission is an extended version of a paper published in the proceedings of "NeSy 2025 – the International Conference on Neurosymbolic Learning and Reasoning" titled ”Grounding Terms from an Ontology for use in Autoformalization: Tokenization is All You Need.” This submission is intended for a Special Issue on NeSy 2025. Notable changes from the previous version of the paper include: * Added table with examples of hallucinated terms produced by fine-tuned LLM. * Addressed potential alternative solutions, such as Retrieval-augmented generation (RAG). * Discussed importance of work to neuro-symbolic reasoning. * Added section outlining evaluation metrics. * Provided discussion and Figure on how synthetic training data was generated. * Re-ordered sections, placing related work after the background section. * Re-titled paper to be more descriptive of content.