The discovery of novel mechanical metamaterials, whose properties are dominated by their engineered structures rather than chemical composition, is a knowledge-intensive and resource-demanding process. To accelerate the design of novel metamaterials, we present MetaScientist, a human-in-the-loop system that integrates advanced AI capabilities with expert oversight with two primary phases: (1) hypothesis generation, where the system performs complex reasoning to generate novel and scientifically sound hypotheses, supported with domain-specific foundation models and inductive biases retrieved from existing literature; (2) 3D structure synthesis, where a 3D structure is synthesized with a novel 3D diffusion model based on the textual hypothesis and refined it with a LLM-based refinement model to achieve better structure properties. At each phase, domain experts iteratively validate the system outputs, and provide feedback and supplementary materials to ensure the alignment of the outputs with scientific principles and human preferences. Through extensive evaluation from human scientists, MetaScientist is able to deliver novel and valid mechanical metamaterial designs that have the potential to be highly impactful in the metamaterial field.

Long-tailed data distributions are prevalent in a variety of domains, including finance, e-commerce, biomedical science, and cyber security. In such scenarios, the performance of machine learning models is often dominated by the head categories, while the learning of tail categories is significantly inadequate. Given abundant studies conducted to alleviate the issue, this work aims to provide a systematic view of long-tailed learning with regard to three pivotal angles: (A1) the characterization of data long-tailedness, (A2) the data complexity of various domains, and (A3) the heterogeneity of emerging tasks. To achieve this, we develop the most comprehensive (to the best of our knowledge) long-tailed learning benchmark named HeroLT, which integrates 13 state-of-the-art algorithms and 6 evaluation metrics on 14 real-world benchmark datasets across 4 tasks from 3 domains. HeroLT with novel angles and extensive experiments (264 in total) enables researchers and practitioners to effectively and fairly evaluate newly proposed methods compared with existing baselines on varying types of datasets. Finally, we conclude by highlighting the significant applications of long-tailed learning and identifying several promising future directions. For accessibility and reproducibility, we open-source our benchmark HeroLT and corresponding results at https://github.com/SSSKJ/HeroLT.

While graph heterophily has been extensively studied in recent years, a fundamental research question largely remains nascent: How and to what extent will graph heterophily affect the prediction performance of graph neural networks (GNNs)? In this paper, we aim to demystify the impact of graph heterophily on GNN spectral filters. Our theoretical results show that it is essential to design adaptive polynomial filters that adapts different degrees of graph heterophily to guarantee the generalization performance of GNNs. Inspired by our theoretical findings, we propose a simple yet powerful GNN named GPatcher by leveraging the MLP-Mixer architectures. Our approach comprises two main components: (1) an adaptive patch extractor function that automatically transforms each node's non-Euclidean graph representations to Euclidean patch representations given different degrees of heterophily, and (2) an efficient patch mixer function that learns salient node representation from both the local context information and the global positional information. Through extensive experiments, the GPatcher model demonstrates outstanding performance on node classification compared with popular homophily GNNs and state-of-the-art heterophily GNNs.

Transferring knowledge across graphs plays a pivotal role in many high-stake domains, ranging from transportation networks to e-commerce networks, from neuroscience to finance. To date, the vast majority of existing works assume both source and target domains are sampled from a universal and stationary distribution. However, many real-world systems are intrinsically dynamic, where the underlying domains are evolving over time. To bridge the gap, we propose to shift the problem to the dynamic setting and ask: given the label-rich source graphs and the label-scarce target graphs observed in previous T timestamps, how can we effectively characterize the evolving domain discrepancy and optimize the generalization performance of the target domain at the incoming T+1 timestamp? To answer the question, for the first time, we propose a generalization bound under the setting of dynamic transfer learning across graphs, which implies the generalization performance is dominated by domain evolution and domain discrepancy between source and target domains. Inspired by the theoretical results, we propose a novel generic framework DyTrans to improve knowledge transferability across dynamic graphs. In particular, we start with a transformer-based temporal encoding module to model temporal information of the evolving domains; then, we further design a dynamic domain unification module to efficiently learn domain-invariant representations across the source and target domains. Finally, extensive experiments on various real-world datasets demonstrate the effectiveness of DyTrans in transferring knowledge from dynamic source domains to dynamic target domains.

Long-tail data distributions are prevalent in many real-world networks, including financial transaction networks, e-commerce networks, and collaboration networks. Despite the success of recent developments, the existing works mainly focus on debiasing the machine learning models via graph augmentation or objective reweighting. However, there is limited literature that provides a theoretical tool to characterize the behaviors of long-tail categories on graphs and understand the generalization performance in real scenarios. To bridge this gap, we propose the first generalization bound for long-tail classification on graphs by formulating the problem in the fashion of multi-task learning, i.e., each task corresponds to the prediction of one particular category. Our theoretical results show that the generalization performance of long-tail classification is dominated by the range of losses across all tasks and the total number of tasks. Building upon the theoretical findings, we propose a novel generic framework Tail2Learn to improve the performance of long-tail categories on graphs. In particular, we start with a hierarchical task grouping module that allows label-limited classes to benefit from the relevant information shared by other classes; then, we further design a balanced contrastive learning module to balance the gradient contributions of head and tail classes. Finally, extensive experiments on various real-world datasets demonstrate the effectiveness of Tail2Learn in capturing long-tail categories on graphs.