提出新方法提升大模型解释性，更准确定位关键词语
Uniform Discretized Integrated Gradients: An effective attribution based method for explaining large language models
- 设计非线性路径插值策略，更好处理词嵌入离散空间
- 在情感分类与问答任务中，三项评估指标均优于现有方法
- 适合需要精准解释的自然语言任务，如模型调试与可信AI
集成梯度是解释深度学习模型的经典方法，通过沿线性路径累积输入特征对输出的梯度来计算特征重要性。然而，该方法在连续特征空间表现良好，对词嵌入等离散空间效果不佳。为解释大语言模型，需在嵌入空间中选择更贴近真实词汇的非线性中间路径。本文提出均匀离散化集成梯度（UDIG），采用新型插值策略，构建适合语言模型的非线性路径以计算归因分数。我们在两类NLP任务上评估：情感分类（使用SST2、IMDb、Rotten Tomatoes数据集）和问答任务（基于微调BERT的SQuAD）。对比三个指标——对数几率、完整性与充分性。结果表明，本方法在几乎所有指标上均优于现有方法。
原文摘要 · Abstract (English)
Integrated Gradients is a well-known technique for explaining deep learning models. It calculates feature importance scores by employing a gradient based approach computing gradients of the model output with respect to input features and accumulating them along a linear path. While this works well for continuous features spaces, it may not be the most optimal way to deal with discrete spaces like word embeddings. For interpreting LLMs (Large Language Models), there exists a need for a non-linear path where intermediate points, whose gradients are to be computed, lie close to actual words in the embedding space. In this paper, we propose a method called Uniform Discretized Integrated Gradients (UDIG) based on a new interpolation strategy where we choose a favorable nonlinear path for computing attribution scores suitable for predictive language models. We evaluate our method on two types of NLP tasks- Sentiment Classification and Question Answering against three metrics viz Log odds, Comprehensiveness and Sufficiency. For sentiment classification, we have used the SST2, IMDb and Rotten Tomatoes datasets for benchmarking and for Question Answering, we have used the fine-tuned BERT model on SQuAD dataset. Our approach outperforms the existing methods in almost all the metrics.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。