Summary of D-separation For Causal Self-explanation, by Wei Liu et al.
D-Separation for Causal Self-Explanation
by Wei Liu, Jun Wang, Haozhao Wang, Ruixuan Li, Zhiying Deng, YuanKai Zhang, Yang Qiu
First submitted to arxiv on: 23 Sep 2023
Categories
- Main: Artificial Intelligence (cs.AI)
- 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 Rationalization is a framework for NLP models that aims to explain the reasoning behind their predictions. The paper proposes a novel approach, called Minimum Conditional Dependence (MCD), which differs from conventional methods like maximum mutual information (MMI). The MCD criterion minimizes the dependence between input features and target labels conditioned on selected rationale candidates. This helps uncover causal rationales by avoiding spurious correlations. Empirically, MCD improves F1 scores by up to 13.7% compared to MMI-based methods. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary Rationalization is a way for AI models to explain their decisions. The problem with current approaches is that they can be fooled by accidental patterns in the data. Instead of trying to fix this, researchers came up with a new idea called Minimum Conditional Dependence (MCD). This method helps the model focus on the real reasons behind its predictions. In tests, MCD did better than previous methods, increasing accuracy. |
Keywords
* Artificial intelligence * Nlp
