Self-Tuning: Instructing LLMs to Effectively Acquire New Knowledge through Self-Teaching
Zhang, Xiaoying, Peng, Baolin, Tian, Ye, Zhou, Jingyan, Zhang, Yipeng, Mi, Haitao, Meng, Helen
–arXiv.org Artificial Intelligence
Large language models (LLMs) often struggle to provide up-to-date information due to their one-time training and the constantly evolving nature of the world. To keep LLMs current, existing approaches typically involve continued pre-training on new documents. However, they frequently face difficulties in extracting stored knowledge. Motivated by the remarkable success of the Feynman Technique in efficient human learning, we introduce Self-Tuning, a learning framework aimed at improving an LLM's ability to effectively acquire new knowledge from raw documents through self-teaching. Specifically, we develop a Self-Teaching strategy that augments the documents with a set of knowledge-intensive tasks created in a self-supervised manner, focusing on three crucial aspects: memorization, comprehension, and self-reflection. In addition, we introduce three Wiki-Newpages-2023-QA datasets to facilitate an in-depth analysis of an LLM's knowledge acquisition ability concerning memorization, extraction, and reasoning. Extensive experimental results on Llama2 family models reveal that Self-Tuning consistently exhibits superior performance across all knowledge acquisition tasks and excels in preserving previous knowledge.
arXiv.org Artificial Intelligence
Jun-15-2024
- Country:
- South America > Chile (0.04)
- North America
- United States
- New York (0.04)
- Arkansas (0.04)
- Mississippi (0.04)
- Colorado (0.04)
- Minnesota > Hennepin County
- Minneapolis (0.14)
- Louisiana > Orleans Parish
- New Orleans (0.04)
- California > San Diego County
- San Diego (0.04)
- Canada > Quebec
- Montreal (0.04)
- United States
- Europe
- Ukraine > Kyiv Oblast
- Kyiv (0.04)
- Spain > Catalonia
- Barcelona Province > Barcelona (0.04)
- Netherlands > North Holland
- Amsterdam (0.04)
- Croatia > Dubrovnik-Neretva County
- Dubrovnik (0.04)
- Ukraine > Kyiv Oblast
- Asia
- China > Hong Kong (0.04)
- Singapore (0.04)
- Middle East
- Jordan (0.04)
- UAE > Abu Dhabi Emirate
- Abu Dhabi (0.04)
- Kazakhstan > Akmola Region
- Astana (0.05)
- Genre:
- Research Report > New Finding (1.00)
- Personal (0.93)
- Industry:
- Media > Film (1.00)
- Education (1.00)
- Law (0.67)
- Leisure & Entertainment > Sports
- Government > Regional Government
- Technology: