This model is trained solely for the purpose of research on Text Embedding Inversion in LLM Security.