Summary of Mitigating Copy Bias in In-context Learning Through Neuron Pruning, by Ameen Ali et al.
Mitigating Copy Bias in In-Context Learning through Neuron Pruning
by Ameen Ali, Lior Wolf, Ivan Titov
First submitted to arxiv on: 2 Oct 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: 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 In this paper, researchers explore a phenomenon known as copying bias in large language models (LLMs). Despite their impressive abilities to learn from few examples, LLMs sometimes prioritize copying answers from provided examples rather than learning underlying patterns. To mitigate this issue, the authors propose a novel and simple method that identifies neurons prioritizing copying over generalization and prunes them. This approach is shown to improve performance across various in-context learning (ICL) tasks, using different architectures such as Transformers and State-Space Models. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Large language models can learn new skills quickly when given a few examples. But sometimes they copy answers from those examples instead of figuring out the underlying rules. The authors of this study want to help LLMs be better at learning by themselves. They developed a simple way to make the models focus more on generalization and less on copying. This method works with different types of language models, and it makes them perform better in various tasks. |
Keywords
» Artificial intelligence » Generalization
