Summary of Explaining Large Language Models Decisions Using Shapley Values, by Behnam Mohammadi
Explaining Large Language Models Decisions Using Shapley Values
by Behnam Mohammadi
First submitted to arxiv on: 29 Mar 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 novel approach presented in this paper uses Shapley values from cooperative game theory to interpret large language model (LLM) behavior and quantify the relative contribution of each prompt component to the model’s output. The authors demonstrate the effectiveness of their method through two applications: a discrete choice experiment and an investigation of cognitive biases. They uncover “token noise” effects, where LLM decisions are disproportionately influenced by tokens providing minimal informative content. This phenomenon raises concerns about the robustness and generalizability of insights obtained from LLMs in the context of human behavior simulation. |
Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper shows that large language models (LLMs) can’t always be trusted to simulate human behavior because they’re influenced by small details in the prompts. The researchers came up with a new way to understand how LLMs work using an idea from game theory. They tested this approach on two real-life examples and found some surprising results. It turns out that LLMs are easily swayed by minor changes in the questions asked, which could lead to unreliable conclusions. |
Keywords
» Artificial intelligence » Large language model » Prompt » Token