Enhancing Visual and Cognitive Intelligence Memory with Fine-Tuning Modeling Based on Artificial Intelligence Large Language Models
Abstract
This study investigates the potential of fine-tuning modeling based on Large Language Models (LLMs) to enhance visual and cognitive intelligence memory. A fine-tuning model was developed using the ChatGPT-4o framework to generate a narrative specifically tailored for a pedagogical audience. This process of guided text generation served as the core of our fine-tuning approach, which yielded the most effective results in creating a foundational narrative. The resulting text was subsequently visualized through detailed prompt engineering on three leading text-to-image AI platforms: DALL-E 3, Imagen 3, and Fooocus.ai. A comparative analysis revealed that the integrated method, initiated with the fine-tuned narrative from the ChatGPT-4o model and visualized with DALL-E 3, produced the most coherent and stylistically consistent outcomes. This synergy, which tightly integrates verbal and visual channels, supports cognitive frameworks like Dual Coding Theory and demonstrates a powerful method for strengthening memory and comprehension. The study highlights significant benefits, such as democratizing creativity and increasing student engagement. However, it also identifies critical risks, including cognitive offloading, the perpetuation of AI-driven biases, and complex copyright issues. In conclusion, this research confirms that fine-tuned AI models are powerful supplementary tools, not teacher replacements. Their effective integration requires a pedagogical shift where assessment focuses on the students’ critical and creative process of guiding the model, rather than on the final AI-generated product.
References
- Gun, M., & Basturk, B. M. (2025). Enhancing visual and cognitive intelligence memory with fine tuning modeling based on artificial intelligence large language models. The Eurasia Proceedings of Science, Technology, Engineering and Mathematics (EPSTEM), 35, 10-28
Details
| Primary Language | English |
|---|---|
| Subjects | Software Engineering (Other) |
| Journal Section | Articles |
| Authors | |
| Early Pub Date | October 20, 2025 |
| Publication Date | October 27, 2025 |
| Submission Date | May 3, 2025 |
| Acceptance Date | June 9, 2025 |
| Published in Issue | Year 2025 Volume: 35 |
