Summary of Can Llms Learn New Concepts Incrementally Without Forgetting?, by Junhao Zheng et al.
Can LLMs Learn New Concepts Incrementally without Forgetting?
by Junhao Zheng, Shengjie Qiu, Qianli Ma
First submitted to arxiv on: 13 Feb 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Computation and Language (cs.CL)
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 This paper explores the incremental learning abilities of Large Language Models (LLMs) by introducing a novel dataset called Concept-1K, which consists of 1,023 recently emerged concepts across diverse domains. The authors aim to answer whether LLMs can learn new concepts incrementally without forgetting like humans. The results show that despite fine-tuning fewer parameters with LoRA, LLMs still suffer from catastrophic forgetting on training data. The paper also investigates the roles of in-context learning, model scale, buffer size, and pretraining in IL performance. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research looks at how well Large Language Models (LLMs) can learn new things without forgetting what they already know. They created a special dataset with lots of small pieces of information about different topics to test this. The results show that LLMs still have trouble learning new things while keeping the old information, even when trying to improve their abilities. The study also looks at how different factors affect how well LLMs do at learning and remembering. |
Keywords
* Artificial intelligence * Fine tuning * Lora * Pretraining
