AI 中文总结
针对神经网络预测厄尔尼诺难以解释的问题，提出平均梯度外积XAI方法，在归因、稳健性和连贯性上优于基线方法，并能揭示物理机制和操控极端事件。
AI 中文摘要
物理科学中一个重要且尚未解决的问题是解释神经网络做出的预测。为解决这一问题，人们提出了多种可解释人工智能（XAI）方法，包括梯度XAI、积分梯度和GradientSHAP。我们根据四个评分来评估这些基线XAI方法：敏感性（XAI模式强烈影响预测）、归因（XAI模式再现相对于基线的预测变化）、稳健性（XAI模式在邻近输入下保持稳定）以及连贯性（XAI模式在空间上平滑）。我们还提出了一种新方法，即平均梯度外积（AGOP）XAI，它利用全局梯度信息为特定输入识别一个重要方向。我们将XAI应用于基于Zebiak-Cane模型数据对厄尔尼诺-南方涛动（ENSO）的神经网络预测。在本文报告的架构和提前期比较中，AGOP XAI取得了最高的归因、稳健性和连贯性得分。其敏感性被梯度XAI超越，后者按定义具有最大敏感性。除了诊断神经网络行为之外，AGOP XAI还能生成关于物理机制的候选假设。该方法突出显示了一个与充放电振荡器物理机制一致的赤道温跃层深度信号，以及一个可能为Zebiak-Cane模型所特有的东南太平洋瓣状结构。最后，我们使用优化扰动来测试AGOP的物理相关性，这些扰动使Zebiak-Cane模型沿AGOP解释坐标移动。此类扰动可以抑制选定的极端事件，或者从接近中性的集合中，在10个月后生成强厄尔尼诺或拉尼娜事件。
英文摘要
An important and unresolved problem in the physical sciences is explaining the predictions made by neural networks. Several explainable artificial intelligence (XAI) methods have been proposed to address this problem, including gradient XAI, Integrated Gradients, and GradientSHAP. We evaluate the baseline XAI methods according to four scores: sensitivity (XAI patterns strongly affect predictions), attribution (XAI patterns reproduce the change in prediction relative to a baseline), robustness (XAI patterns remain stable for nearby inputs), and coherence (XAI patterns are spatially smooth). We also introduce a new method, average gradient outer product (AGOP) XAI, that uses global gradient information to identify an important direction for a specific input. We apply XAI to neural network predictions of the El Niño-Southern Oscillation (ENSO) based on data from the Zebiak-Cane model. AGOP XAI achieves the highest attribution, robustness, and coherence scores in the architecture and lead-time comparisons reported here. Its sensitivity is surpassed by gradient XAI, which is maximally sensitive by definition. Beyond diagnosing neural-network behavior, AGOP XAI can generate candidate hypotheses about physical mechanisms. The method highlights an equatorial thermocline-depth signal consistent with recharge oscillator physics, together with a southeastern-Pacific lobe that may be specific to the Zebiak-Cane model. Finally, we test the physical relevance of AGOP using optimized perturbations that move the Zebiak-Cane model along AGOP explanation coordinates. Such perturbations can suppress the selected extreme events or, from a near-neutral ensemble, generate strong El Niño or La Niña events 10 months later.
Comments21 pages, 12 figures, 1 table. Software: https://github.com/rjwebber/agop-xai/releases/tag/v1.0.0. Processed data, trained models, and reproducibility artifacts: https://doi.org/10.5281/zenodo.23053004
