比较神经网络逻辑编码，提升可解释性效率
Comparing Neural Network Encodings for Logic-based Explainability
- 对比两种神经网络逻辑编码方法，优化可解释性构建
- 新编码减少变量与约束，构建时间快18%，总耗时快16%
- 适合关注模型可解释性与计算效率的研究者
为人工神经网络（ANN）输出提供解释在关键系统、数据保护法规及对抗样本处理中至关重要。基于逻辑的方法可提供具有正确性保证的解释，但面临可扩展性挑战。因此，有必要比较不同将神经网络编码为逻辑约束的方法，这些方法用于逻辑可解释性。本文比较了两种编码：一种已在文献中用于解释，另一种将被适配到本研究的可解释性场景。后者使用的变量和约束更少，可能提升效率。实验显示，解释计算时间相似，但新编码在构建逻辑约束上最多提升18%，整体时间最多提升16%。
原文摘要 · Abstract (English)
Providing explanations for the outputs of artificial neural networks (ANNs) is crucial in many contexts, such as critical systems, data protection laws and handling adversarial examples. Logic-based methods can offer explanations with correctness guarantees, but face scalability challenges. Due to these issues, it is necessary to compare different encodings of ANNs into logical constraints, which are used in logic-based explainability. This work compares two encodings of ANNs: one has been used in the literature to provide explanations, while the other will be adapted for our context of explainability. Additionally, the second encoding uses fewer variables and constraints, thus, potentially enhancing efficiency. Experiments showed similar running times for computing explanations, but the adapted encoding performed up to 18\% better in building logical constraints and up to 16\% better in overall time.
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。