https://arxiv.org/abs/2312.05934

0:00 Comparing Finetuning and Retrieval Augmented Generation
0:34 Using LLMs for Specialized Domains
1:13 Finetuning vs Incontext Learning Techniques
2:23 Causes of LLM Factual Errors and Hallucinations
3:50 Constructing the Experiment Dataset
4:45 Models Tested and Accuracy Comparison
5:51 RAG Outperforms Finetuning Across Models
6:20 Why RAG Performs Better Than Finetuning
7:01 Caveats and Open Questions
7:39 Conclusion and Wrapup


Video explaining MMLU and other benchmarks:    • LLM benchmarks