Ge Cheng

Libin Liang, Ye Tian, Ge Cheng

Study of neural networks with infinite width is important for better understanding of the neural network in practical application. In this work, we derive the equivalence of the deep, infinite-width maxout network and the Gaussian process (GP) and characterize the maxout kernel with a compositional structure. Moreover, we build up the connection between our deep maxout network kernel and deep neural network kernels. We also give an efficient numerical implementation of our kernel which can be adapted to any maxout rank. Numerical results show that doing Bayesian inference based on the deep maxout network kernel can lead to competitive results compared with their finite-width counterparts and deep neural network kernels. This enlightens us that the maxout activation may also be incorporated into other infinite-width neural network structures such as the convolutional neural network (CNN).

Ge Cheng, Shuo Wang, Yun Zhang

Contrastive learning has emerged as a cornerstone of unsupervised representation learning across vision, language, and graph domains, with InfoNCE as its dominant objective. Despite its empirical success, the theoretical underpinnings of InfoNCE remain limited. In this work, we introduce an explicit feature space to model augmented views of samples and a transition probability matrix to capture data augmentation dynamics. We demonstrate that InfoNCE optimizes the probability of two views sharing the same source toward a constant target defined by this matrix, naturally inducing feature clustering in the representation space. Leveraging this insight, we propose Scaled Convergence InfoNCE (SC-InfoNCE), a novel loss function that introduces a tunable convergence target to flexibly control feature similarity alignment. By scaling the target matrix, SC-InfoNCE enables flexible control over feature similarity alignment, allowing the training objective to better match the statistical properties of downstream data. Experiments on benchmark datasets, including image, graph, and text tasks, show that SC-InfoNCE consistently achieves strong and reliable performance across diverse domains.

Shuo Wang, Ge Cheng, Yun Zhang

Message Passing Neural Networks (MPNNs) model local relations effectively but struggle to propagate information over long distances. Graph Transformers (GTs) mitigate this via global self-attention, yet their quadratic cost in the number of nodes limits scalability. We propose Graph Elimination Networks (GENs), an MPNN variant that approximates GT-like long-range modeling while maintaining high efficiency. GENs combine edge-wise and hop-wise self-attention in parallel; their multiplicative composition yields an attention kernel separable across edge and hop factors within a bounded K-hop receptive field. To enable hop-wise attention, we introduce the Graph Elimination Algorithm (GEA), which prevents double counting across hops, ensuring that each round injects the k-hop incremental contribution exactly once. Taking differences between successive rounds recovers the k-hop increment and yields disentangled multi-hop features as inputs for hop-wise attention. This preserves clearer structural distinctions across hop distances and enables more faithful modeling of pairwise dependencies between distant nodes within the K-hop neighborhood. On the Long-Range Graph Benchmark (LRGB), GENs outperform strong MPNN baselines by 7.7 and 6.0 percentage points (pp) on PascalVOC-SP and COCO-SP, and achieve performance on par with or better than state-of-the-art Graph Transformers. On OGBN-Products, GENs support full-batch training/inference, while sparse-attention baselines like Exphormer struggle with memory limits under comparable budgets, highlighting GENs as a practical alternative for large, sparse graphs.

Yuansheng Liu, Hua Fan, Eric Yong Xie et al.

Since John von Neumann suggested utilizing Logistic map as a random number generator in 1947, a great number of encryption schemes based on Logistic map and/or its variants have been proposed. This paper re-evaluates the security of an image cipher based on transformed logistic maps and proves that the image cipher can be deciphered efficiently under two different conditions: 1) two pairs of known plain-images and the corresponding cipher-images with computational complexity of $O(2^{18}+L)$; 2) two pairs of chosen plain-images and the corresponding cipher-images with computational complexity of $O(L)$, where $L$ is the number of pixels in the plain-image. In contrast, the required condition in the previous deciphering method is eighty-seven pairs of chosen plain-images and the corresponding cipher-images with computational complexity of $O(2^{7}+L)$. In addition, three other security flaws existing in most Logistic-map-based ciphers are also reported.

Chengqing Li, Tao Xie, Qi Liu et al.

Chaotic behavior arises from very simple non-linear dynamical equation of logistic map which makes it was used often in designing chaotic image encryption schemes. However, some properties of chaotic maps can also facilitate cryptanalysis especially when they are implemented in digital domain. Utilizing stable distribution of the chaotic states generated by iterating the logistic map, this paper presents a typical example to show insecurity of an image encryption scheme using chaotic logistic map. This work will push encryption and chaos be combined in a more effective way.