Summary of Unveiling Imitation Learning: Exploring the Impact Of Data Falsity to Large Language Model, by Hyunsoo Cho
Unveiling Imitation Learning: Exploring the Impact of Data Falsity to Large Language Model
by Hyunsoo Cho
First submitted to arxiv on: 15 Apr 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
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 The paper explores the impact of noisy data on open-source language models through instruction tuning. It introduces the Falsity-Controllable (FACO) dataset, which allows for manual control of falsity ratios. The authors find a strong correlation between factuality and benchmark scores, and that training LLMs with false instructions leads to generation of fake answers. Additionally, they show that restoring original performance is possible but doesn’t reach full potential. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at how noisy data affects language models. Noisy data can be bad because it’s not accurate, which makes the model learn things that aren’t true. The authors created a special dataset called FACO to study this problem. They found that if you train the model with fake instructions, it will start giving false answers even when it knows the correct answer. This is a problem because it means the model isn’t learning from its mistakes and can’t improve. |
Keywords
» Artificial intelligence » Instruction tuning
