Recent advances in generative models have demonstrated remarkable capabilities in producing high-quality images, but their reliance on large-scale unlabeled data has raised significant safety and copyright concerns. Efforts to address these issues by erasing unwanted concepts have shown promise. However, many existing erasure methods involve excessive modifications that compromise the overall utility of the model. In this work, we address these issues by formulating a novel minimalist concept erasure objective based *only* on the distributional distance of final generation outputs. Building on our formulation, we derive a tractable loss for differentiable optimization that leverages backpropagation through all generation steps in an end-to-end manner. We also conduct extensive analysis to show theoretical connections with other models and methods. To improve the robustness of the erasure, we incorporate neuron masking as an alternative to model fine-tuning. Empirical evaluations on state-of-the-art flow-matching models demonstrate that our method robustly erases concepts without degrading overall model performance, paving the way for safer and more responsible generative models.

We investigate Graph Neural Networks (GNNs) on heterophilous graphs for node classification. To address the scarcity of useful local information in heterophilous neighborhood, it is often essential to explore global interactions. However, many existing methods in this endeavor are computationally expensive and may suffer from issues like oversquashing. In addition, earlier studies show that GNNs can be outperformed by Multi-Layer Perceptrons on heterophilous graphs, indicating insufficient exploitation of node feature information. To address these limitations, we propose Prototype Mediated GNN (PM-GNN), a novel framework which efficiently captures global feature information using class prototypes. PM-GNN learns multiple class prototypes for each class from raw node features with a soft k-means clustering mechanism. These prototypes are then transferred onto node embeddings via explicit message passing, bypassing local neighborhoods and mitigating oversquashing. PM-GNN can scale to large graphs, outperforming strong baselines on multiple heterophilous datasets.

This paper addresses the challenge of object-centric lay-out generation under spatial constraints, seen in multi-ple domains including floorplan design process. The de-sign process typically involves specifying a set of spa-tial constraints that include object attributes like size and inter-object relations such as relative positioning. Existing works, which typically represent objects as single nodes, lack the granularity to accurately model complex interactions between objects. For instance, often only certain parts of an object, like a room's right wall, interact with adjacent objects. To address this gap, we introduce a factor graph based approach with four latent variable nodes for each room, and a factor node for each constraint. The factor nodes represent dependencies among the variables to which they are connected, effectively capturing constraints that are potentially of a higher order. We then develop message-passing on the bipartite graph, forming a factor graph neu-ral network that is trained to produce a floorplan that aligns with the desired requirements. Our approach is simple and generates layouts faithful to the user requirements, demon-strated by a large improvement in IOU scores over existing methods. Additionally, our approach, being inferential and accurate, is well-suited to the practical human-in-the-loop design process where specifications evolve iteratively, offering a practical and powerful tool for AI-guided design.

Efficient news exploration is crucial in real-world applications, particularly within the financial sector, where numerous control and risk assessment tasks rely on the analysis of public news reports. The current processes in this domain predominantly rely on manual efforts, often involving keyword-based searches and the compilation of extensive keyword lists. In this paper, we introduce NCEXPLORER, a framework designed with OLAP-like operations to enhance the news exploration experience. NCEXPLORER empowers users to use roll-up operations for a broader content overview and drill-down operations for detailed insights. These operations are achieved through integration with external knowledge graphs (KGs), encompassing both fact-based and ontology-based structures. This integration significantly augments exploration capabilities, offering a more comprehensive and efficient approach to unveiling the underlying structures and nuances embedded in news content. Extensive empirical studies through master-qualified evaluators on Amazon Mechanical Turk demonstrate NCEXPLORER'S superiority over existing state-of-the-art news search methodologies across an array of topic domains, using real-world news datasets.

Existing neural constructive solvers for routing problems have predominantly employed transformer architectures, conceptualizing the route construction as a set-to-sequence learning task. However, their efficacy has primarily been demonstrated on entirely random problem instances that inadequately capture real-world scenarios. In this paper, we introduce realistic Traveling Salesman Problem (TSP) scenarios relevant to industrial settings and derive the following insights: (1) The optimal next node (or city) to visit often lies within proximity to the current node, suggesting the potential benefits of biasing choices based on current locations. (2) Effectively solving the TSP requires robust tracking of unvisited nodes and warrants succinct grouping strategies. Building upon these insights, we propose integrating a learnable choice layer inspired by Hypernetworks to prioritize choices based on the current location, and a learnable approximate clustering algorithm inspired by the Expectation-Maximization algorithm to facilitate grouping the unvisited cities. Together, these two contributions form a hierarchical approach towards solving the realistic TSP by considering both immediate local neighbourhoods and learning an intermediate set of node representations. Our hierarchical approach yields superior performance compared to both classical and recent transformer models, showcasing the efficacy of the key designs.

In this paper, we propose a new class of distributed joint source-channel coding (DJSCC) methods, namely triple polar codes (T-PC), for transmitting a pair of correlated binary sources over noisy channels. In the T-PC structure, one source is protected by a systematic polar code (SPC), and the other source is encoded into a double polar code (D-PC) word. Following this, we prove the T-PC approaches the corner point of the achievable rate-region of DJSCC. We further propose a distributed joint source-channel decoding algorithm, which involves two components: a cyclic redundancy check (CRC) aided successive cancellation list (CA-SCL) decoding of the SPC and a joint successive cancellation list (J-SCL) decoding of the D-PC. The CA-SCL and J-SCL decoding procedures alternately generate hard-decisions of sources which are iteratively exchanged as the side information and result in superior performance compared with the state-of-the-art polar code based DJSCC scheme.

A Graph Convolutional Network (GCN) stacks several layers and in each layer performs a PROPagation operation~(PROP) and a TRANsformation operation~(TRAN) for learning node representations over graph-structured data. Though powerful, GCNs tend to suffer performance drop when the model gets deep. Previous works focus on PROPs to study and mitigate this issue, but the role of TRANs is barely investigated. In this work, we study performance degradation of GCNs by experimentally examining how stacking only TRANs or PROPs works. We find that TRANs contribute significantly, or even more than PROPs, to declining performance, and moreover that they tend to amplify node-wise feature variance in GCNs, causing variance inflammation that we identify as a key factor for causing performance drop. Motivated by such observations, we propose a variance-controlling technique termed Node Normalization (NodeNorm), which scales each node's features using its own standard deviation. Experimental results validate the effectiveness of NodeNorm on addressing performance degradation of GCNs. Specifically, it enables deep GCNs to outperform shallow ones in cases where deep models are needed, and to achieve comparable results with shallow ones on 6 benchmark datasets. NodeNorm is a generic plug-in and can well generalize to other GNN architectures. Code is publicly available at https://github.com/miafei/NodeNorm.

A good quantization scheme plays a tremendously important role in the hardware implementation of the decoder for polar codes. In this paper, a non-uniform quantization scheme for cyclic redundancy check aided successive-cancellation list (CA-SCL) decoder of polar codes is proposed to save memory resources. In the proposed non-uniform quantization scheme, we introduce a simple compression function to decrease the truncation threshold so that the soft information can be quantized with fewer quantization bits. The scaling factor in the compression function is a negative integer power of two, which can be efficiently implemented by shift operations. Then a compressed-domain CA-SCL decoder is derived, alleviating the need to decompress samples one-by-one. Simulation results show that on the condition of approaching floating-point performance, the internal log-likelihood ratio (LLR) memory resource consumption can be saved by 20% compared with uniform quantization.