Summary of Knowtuning: Knowledge-aware Fine-tuning For Large Language Models, by Yougang Lyu et al.
KnowTuning: Knowledge-aware Fine-tuning for Large Language Models
by Yougang Lyu, Lingyong Yan, Shuaiqiang Wang, Haibo Shi, Dawei Yin, Pengjie Ren, Zhumin Chen, Maarten de Rijke, Zhaochun Ren
First submitted to arxiv on: 17 Feb 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary Large language models have shown remarkable success in natural language processing tasks but struggle to leverage knowledge for knowledge-intensive tasks, often generating incomplete, non-factual, or illogical answers. This limitation stems from inadequate knowledge awareness during vanilla fine-tuning. To address this issue, we propose a knowledge-aware fine-tuning (KnowTuning) method that improves fine-grained and coarse-grained knowledge awareness of large language models (LLMs). Our approach consists of two stages: a fine-grained knowledge augmentation stage to train LLMs to identify difficult fine-grained knowledge in answers and a coarse-grained knowledge comparison stage to distinguish between reliable and unreliable knowledge, evaluated across three aspects: completeness, factuality, and logicality. Extensive experiments on both generic and medical question answering datasets demonstrate the effectiveness of KnowTuning through automatic and human evaluations, across various sizes of LLMs. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Large language models are super smart at understanding language but struggle to use this knowledge to answer tricky questions. They often give incomplete or wrong answers because they don’t really understand what they’re talking about. To fix this problem, we developed a new way to train these models called KnowTuning. It helps them learn to recognize when their answers are correct or not. We tested KnowTuning on lots of questions and it worked really well! Our results show that KnowTuning makes the models better at giving accurate answers. |
Keywords
» Artificial intelligence » Fine tuning » Natural language processing » Question answering
