High-level automation is increasingly critical in AI, driven by rapid advances in large language models (LLMs) and AI agents. However, LLMs, despite their general reasoning power, struggle significantly in specialized, data-sensitive tasks such as designing Graph Neural Networks (GNNs). This difficulty arises from (1) the inherent knowledge gaps in modeling the intricate, varying relationships between graph properties and suitable architectures and (2) the external noise from misleading descriptive inputs, often resulting in generic or even misleading model suggestions. Achieving proficiency in designing data-aware models—defined as the meta-level capability to systematically accumulate, interpret, and apply data-specific design knowledge—remains challenging for existing automated approaches, due to their inefficient construction and application of meta-knowledge. To achieve meta-level proficiency, we propose DesiGNN, a knowledge-centered framework that systematically converts past model design experience into structured, fine-grained knowledge priors well-suited for meta-learning with LLMs. To account for the inherent variability and external noise, DesiGNN aligns empirical property filtering from extensive benchmarks with adaptive elicitation of literature insights via LLMs. By constructing a solid meta-knowledge between unseen graph understanding and known effective architecture patterns, DesiGNN can deliver top-5.77% initial model proposals for unseen datasets within seconds and achieve consistently superior performance with minimal search cost compared to baselines.

Citation classification, which identifies the intention behind academic citations, is pivotal for scholarly analysis. Previous works suggest fine-tuning pretrained language models (PLMs) on citation classification datasets, reaping the reward of the linguistic knowledge they gained during pretraining. However, directly fine-tuning for citation classification is challenging due to labeled data scarcity, contextual noise, and spurious keyphrase correlations. In this paper, we present a novel framework, Citss, that adapts the PLMs to overcome these challenges. Citss introduces self-supervised contrastive learning to alleviate data scarcity, and is equipped with two specialized strategies to obtain the contrastive pairs: sentence-level cropping, which enhances focus on target citations within long contexts, and keyphrase perturbation, which mitigates reliance on specific keyphrases. Compared with previous works that are only designed for encoder-based PLMs, Citss is carefully developed to be compatible with both encoder-based PLMs and decoder-based LLMs, to embrace the benefits of enlarged pretraining. Experiments with three benchmark datasets with both encoder-based PLMs and decoder-based LLMs demonstrate our superiority compared to the previous state of the art. Our code is available at: github.com/LITONG99/Citss

Annotation through crowdsourcing draws incremental attention, which relies on an effective selection scheme given a pool of workers. Existing methods propose to select workers based on their performance on tasks with ground truth, while two important points are missed. 1) The historical performances of workers in other tasks. In real-world scenarios, workers need to solve a new task whose correlation with previous tasks is not well-known before the training, which is called cross-domain. 2) The dynamic worker performance as workers will learn from the ground truth. In this paper, we consider both factors in designing an allocation scheme named cross-domain-aware worker selection with training approach. Our approach proposes two estimation modules to both statistically analyze the cross-domain correlation and simulate the learning gain of workers dynamically. A framework with a theoretical analysis of the worker elimination process is given. To validate the effectiveness of our methods, we collect two novel real-world datasets and generate synthetic datasets. The experiment results show that our method outperforms the baselines on both real-world and synthetic datasets.

Recently, emergence has received widespread attention from the research community along with the success of large-scale models.Different from the literature, we hypothesize a key factor that promotes the performance during the increase of scale: the reduction of monosemantic neurons that can only form one-to-one correlations with specific features.Monosemantic neurons tend to be sparser and have negative impacts on the performance in large models.Inspired by this insight, we propose an intuitive idea to identify monosemantic neurons and inhibit them.However, achieving this goal is a non-trivial task as there is no unified quantitative evaluation metric and simply banning monosemantic neurons does not promote polysemanticity in neural networks.Therefore, we first propose a new metric to measure the monosemanticity of neurons with the guarantee of efficiency for online computation, then introduce a theoretically supported method to suppress monosemantic neurons and proactively promote the ratios of polysemantic neurons in training neural networks.We validate our conjecture that monosemanticity brings about performance change at different model scales on a variety of neural networks and benchmark datasets in different areas, including language, image, and physics simulation tasks.Further experiments validate our analysis and theory regarding the inhibition of monosemanticity.

Precipitation forecasting, modeled as an important chaotic system in earth system science, is not explicitly solved with theory-driven models. In recent years, deep learning models have achieved great success in various applications including rainfall prediction. However, these models work in an image processing manner regardless of the nature of a physical system. We found that the non-linearity relationships learned by deep learning models, which mostly rely on the activation functions, are commonly weighted piecewise continuous functions with bounded first-order derivatives. In contrast, the polynomial is one of the most widely used classes of functions for theory-driven models, applied to numerical approximation, dynamic system modeling, etc.. Researchers started to use the polynomial activation functions (Pacs in short) for neural networks from the 1990s. In recent years, with bloomed researches that apply deep learning to scientific problems, it is weird that such a powerful class of basis functions is rarely used. In this paper, we investigate it and argue that, even though polynomials are good at information extraction, it is too fragile to train stably. We finally solve its serious data flow explosion problem with Chebyshev polynomials and prepended normalization, which enables networks to go deep with Pacs. To enhance the robustness of training, a normalization called Range Norm is further proposed. Performance on synthetic dataset and summer precipitation prediction task validates the necessity of such a class of activation functions to simulate complex physical mechanisms. The new tool for deep learning enlightens a new way of automatic theoretical physics analysis.

Nowadays, ridesharing becomes a popular commuting mode. Dynamically arriving riders post their origins and destinations, then the platform assigns drivers to serve them. In ridesharing, different groups of riders can be served by one driver if their trips can share common routes. Recently, many ridesharing companies (e.g., Didi and Uber) further propose a new mode, namely "ridesharing with meeting points". Specifically, with a short walking distance but less payment, riders can be picked up and dropped off around their origins and destinations, respectively. In addition, meeting points enables more flexible routing for drivers, which can potentially improve the global profit of the system. In this paper, we first formally define the Meeting-Point-based Online Ridesharing Problem (MORP). We prove that MORP is NP-hard and there is no polynomial-time deterministic algorithm with a constant competitive ratio for it. We notice that a structure of vertex set, k -skip cover, fits well to the MORP. k -skip cover tends to find the vertices (meeting points) that are convenient for riders and drivers to come and go. With meeting points, MORP tends to serve more riders with these convenient vertices. Based on the idea, we introduce a convenience-based meeting point candidates selection algorithm. We further propose a hierarchical meeting-point oriented graph (HMPO graph), which ranks vertices for assignment effectiveness and constructs k -skip cover to accelerate the whole assignment process. Finally, we utilize the merits of k -skip cover points for ridesharing and propose a novel algorithm, namely SMDB, to solve MORP. Extensive experiments on real and synthetic datasets validate the effectiveness and efficiency of our algorithms.

In this paper, we study the privacy-preserving task assignment problem in spatial crowdsourcing, where the locations of both workers and tasks, prior to their release to the server, are perturbed with Geo-Indistinguishability (a differential privacy notion for location-based systems). Different from the previously studied online setting, where each task is assigned immediately upon arrival, we target the batch-based setting, where the server maximizes the number of successfully assigned tasks after a batch of tasks arrive. To achieve this goal, we propose the k-Switch solution, which first divides the workers into small groups based on the perturbed distance between workers/tasks, and then utilizes Homomorphic Encryption (HE) based secure computation to enhance the task assignment. Furthermore, we expedite HE-based computation by limiting the size of the small groups under k. Extensive experiments demonstrate that, in terms of the number of successfully assigned tasks, the k-Switch solution improves batch-based baselines by 5.9X and the existing online solution by 1.74X, with no privacy leak.

The dramatic development of shared mobility in food delivery, ridesharing, and crowdsourced parcel delivery has drawn great concerns. Specifically, shared mobility refers to transferring or delivering more than one passenger/package together when their traveling routes have common sub-routes or can be shared. A core problem for shared mobility is to plan a route for each driver to fulfill the requests arriving dynamically with given objectives. Previous studies greedily and incrementally insert each newly coming request to the most suitable worker with a minimum travel cost increase, which only considers the current situation and thus not optimal. In this paper, we propose a demand-aware route planning (DARP) for shared mobility services. Based on prediction, DARP tends to make optimal route planning with more information about requests in the future. We prove that the DARP problem is NP-hard, and further show that there is no polynomial-time deterministic algorithm with a constant competitive ratio for the DARP problem unless P=NP. Hence, we devise an approximation algorithm to realize the insertion operation for our goal. With the insertion algorithm, we devise a prediction based solution for the DARP problem. Extensive experiment results on real datasets validate the effectiveness and efficiency of our technique.