Deep Neural Networks (DNNs) are powerful algorithms that have been proven capable of extracting non-Gaussian information from weak lensing (WL) data sets. Understanding which features in the data determine the output of these nested, non-linear algorithms is an important but challenging task. We analyze a DNN that has been found in previous work to accurately recover cosmological parameters in simulated maps of the WL convergence ($\kappa$). We derive constraints on the cosmological parameter pair $(\Omega_m,\sigma_8)$ from a combination of three commonly used WL statistics (power spectrum, lensing peaks, and Minkowski functionals), using ray-traced simulated $\kappa$ maps. We show that the network can improve the inferred parameter constraints relative to this combination by $20\%$ even in the presence of realistic levels of shape noise. We apply a series of well established saliency methods to interpret the DNN and find that the most relevant pixels are those with extreme $\kappa$ values. For noiseless maps, regions with negative $\kappa$ account for $86-69\%$ of the attribution of the DNN output, defined as the square of the saliency in input space. In the presence of shape nose, the attribution concentrates in high convergence regions, with $36-68\%$ of the attribution in regions with $\kappa > 3 \sigma_{\kappa}$.

中文翻译：

---

解释弱透镜的深度学习模型

深度神经网络 (DNN) 是强大的算法，已被证明能够从弱透镜 (WL) 数据集中提取非高斯信息。了解数据中的哪些特征决定了这些嵌套非线性算法的输出是一项重要但具有挑战性的任务。我们分析了在先前工作中发现的 DNN，以准确恢复 WL 收敛的模拟图中的宇宙学参数（$\kappa$）。我们使用光线追踪模拟的 $\kappa$ 地图，从三个常用的 WL 统计（功率谱、透镜峰和 Minkowski 泛函）的组合中得出对宇宙学参数对 $(\Omega_m,\sigma_8)$ 的约束。我们表明，即使存在真实的形状噪声水平，网络也可以将相对于这种组合的推断参数约束提高 20% 美元。我们应用一系列完善的显着性方法来解释 DNN，并发现最相关的像素是那些具有极端 $\kappa$ 值的像素。对于无噪声地图，负值 $\kappa$ 的区域占 DNN 输出属性的 $86-69\%$，定义为输入空间中显着性的平方。在存在形状鼻子的情况下，归因集中在高收敛区域，$\kappa > 3 \sigma_{\kappa}$ 区域的归因有$36-68\%$。