Evaluation of Embedding-Based and Generative Methods for LLM-Driven Document Classification: Opportunities and Challenges

1. 1Evaluate the performance of embedding-based models for document classification
2. 2Compare the results with generative Vision-Language Models (VLMs) like Qwen2.5-VL
3. 3Investigate the impact of Chain-of-Thought (CoT) prompting on zero-shot accuracy
4. 4Analyze the trade-offs between model accuracy, stability, and computational cost

AI engineers and researchers on a team benefit from this study as it provides insights into the trade-offs between model accuracy, stability, and computational cost, while data scientists can apply these findings to improve document classification tasks

💡 Generative Vision-Language Models with Chain-of-Thought prompting outperform embedding-based models in document classification tasks

💡 Generative VLMs achieve 82% zero-shot accuracy for document classification with CoT prompting