Simultaneously achieving robust classification and high-fidelity generative modeling within a single framework presents a significant challenge. Hybrid approaches, such as Joint Energy-Based Models (JEM), interpret classifiers as EBMs but are often limited by the instability and poor sample quality inherent in Stochastic Gradient Langevin Dynamics (SGLD)-based training. We address these limitations by proposing a novel training framework that integrates adversarial training (AT) principles for both discriminative robustness and stable generative learning. The proposed method introduces three key innovations: (1) the replacement of SGLD-based JEM learning with a stable, AT-based approach that optimizes the energy function by discriminating between real data and Projected Gradient Descent (PGD)-generated contrastive samples using the BCE loss; (2) synergistic adversarial training for the discriminative component that enhances classification robustness while eliminating the need for explicit gradient penalties; and (3) a two-stage training strategy that addresses normalization-related instabilities and enables leveraging pretrained robust classifiers, generalizing effectively across diverse architectures. Experiments on CIFAR-10/100 and ImageNet demonstrate that our approach: (1) is the first EBM-based hybrid to scale to high-resolution datasets with high training stability, simultaneously achieving state-of-the-art discriminative and generative performance on ImageNet 256$\times$256; (2) uniquely combines generative quality with adversarial robustness, enabling critical applications like robust counterfactual explanations; and (3) functions as a competitive standalone generative model, matching the generative quality of autoregressive methods (VAR-d16) and surpassing diffusion models while offering unique versatility.

翻译：在单一框架内同时实现鲁棒的分类和高保真的生成建模是一项重大挑战。混合方法，如联合能量基模型（JEM），将分类器解释为能量基模型，但通常受限于基于随机梯度朗之万动力学（SGLD）训练固有的不稳定性和较差的样本质量。我们通过提出一种新颖的训练框架来解决这些局限性，该框架集成了对抗训练（AT）原理，旨在同时实现判别鲁棒性和稳定的生成学习。所提出的方法引入了三项关键创新：（1）用稳定的、基于对抗训练的方法取代基于SGLD的JEM学习，该方法通过使用二元交叉熵损失来区分真实数据与投影梯度下降（PGD）生成的对比样本，从而优化能量函数；（2）针对判别组件的协同对抗训练，在增强分类鲁棒性的同时，消除了对显式梯度惩罚的需求；（3）一种两阶段训练策略，解决了与归一化相关的不稳定性，并能够利用预训练的鲁棒分类器，从而在不同架构上实现有效泛化。在CIFAR-10/100和ImageNet上的实验表明，我们的方法：（1）是首个能够扩展到高分辨率数据集且具有高训练稳定性的基于能量基模型的混合方法，同时在ImageNet 256$\times$256上实现了最先进的判别和生成性能；（2）独特地将生成质量与对抗鲁棒性相结合，使得诸如鲁棒反事实解释等关键应用成为可能；（3）作为一个具有竞争力的独立生成模型，其生成质量与自回归方法（VAR-d16）相当，并超越了扩散模型，同时提供了独特的通用性。