CATP: Contextually Adaptive Token Pruning for Efficient and Enhanced Multimodal In-Context Learning

Modern large vision-language models (LVLMs) convert each input image into a large set of tokens, far outnumbering the text tokens. Although this improves visual perception, it introduces severe image token redundancy. Because image tokens carry sparse information, many add little to reasoning, yet greatly increase inference cost. The emerging image token pruning methods tackle this issue by identifying the most important tokens and discarding the rest. These methods can raise efficiency with only modest performance loss. However, most of them only consider single-image tasks and overlook multimodal in-context learning (ICL), where redundancy is greater and efficiency is more critical. Redundant tokens weaken the advantage of multimodal ICL for rapid domain adaptation and cause unstable performance. Applying existing pruning methods in this setting leads to large accuracy drops, exposing a clear gap and the need for new techniques. Thus, we propose Contextually Adaptive Token Pruning (CATP), a training-free pruning method targeted at multimodal ICL. CATP consists of two stages that perform progressive pruning to fully account for the complex cross-modal interactions in the input sequence. After removing 77.8\% of the image tokens, CATP produces an average performance gain of 0.6\% over the vanilla model on four LVLMs and eight benchmarks, exceeding all baselines remarkably. Meanwhile, it effectively improves efficiency by achieving an average reduction of 10.78\% in inference latency. CATP enhances the practical value of multimodal ICL and lays the groundwork for future progress in interleaved image-text scenarios.

Graph Synthetic Out-of-Distribution Exposure with Large Language Models

Out-of-distribution (OOD) detection in graphs is critical for ensuring model robustness in open-world and safety-sensitive applications. Existing approaches to graph OOD detection typically involve training an in-distribution (ID) classifier using only ID data, followed by the application of post-hoc OOD scoring techniques. Although OOD exposure - introducing auxiliary OOD samples during training - has proven to be an effective strategy for enhancing detection performance, current methods in the graph domain generally assume access to a set of real OOD nodes. This assumption, however, is often impractical due to the difficulty and cost of acquiring representative OOD samples. In this paper, we introduce GOE-LLM, a novel framework that leverages Large Language Models (LLMs) for OOD exposure in graph OOD detection without requiring real OOD nodes. GOE-LLM introduces two pipelines: (1) identifying pseudo-OOD nodes from the initially unlabeled graph using zero-shot LLM annotations, and (2) generating semantically informative synthetic OOD nodes via LLM-prompted text generation. These pseudo-OOD nodes are then used to regularize the training of the ID classifier for improved OOD awareness. We evaluate our approach across multiple benchmark datasets, showing that GOE-LLM significantly outperforms state-of-the-art graph OOD detection methods that do not use OOD exposure and achieves comparable performance to those relying on real OOD data.

GLIP-OOD: Zero-Shot Graph OOD Detection with Foundation Model

Out-of-distribution (OOD) detection is critical for ensuring the safety and reliability of machine learning systems, particularly in dynamic and open-world environments. In the vision and text domains, zero-shot OOD detection - which requires no training on in-distribution (ID) data - has made significant progress through the use of large-scale pretrained models such as vision-language models (VLMs) and large language models (LLMs). However, zero-shot OOD detection in graph-structured data remains largely unexplored, primarily due to the challenges posed by complex relational structures and the absence of powerful, large-scale pretrained models for graphs. In this work, we take the first step toward enabling zero-shot graph OOD detection by leveraging a graph foundation model (GFM). We show that, when provided only with class label names, the GFM can perform OOD detection without any node-level supervision - outperforming existing supervised methods across multiple datasets. To address the more practical setting where OOD label names are unavailable, we introduce GLIP-OOD, a novel framework that employs LLMs to generate semantically informative pseudo-OOD labels from unlabeled data. These labels enable the GFM to capture nuanced semantic boundaries between ID and OOD classes and perform fine-grained OOD detection - without requiring any labeled nodes. Our approach is the first to enable node-level graph OOD detection in a fully zero-shot setting, and achieves state-of-the-art performance on four benchmark text-attributed graph datasets.

Few-Shot Graph Out-of-Distribution Detection with LLMs

Existing methods for graph out-of-distribution (OOD) detection typically depend on training graph neural network (GNN) classifiers using a substantial amount of labeled in-distribution (ID) data. However, acquiring high-quality labeled nodes in text-attributed graphs (TAGs) is challenging and costly due to their complex textual and structural characteristics. Large language models (LLMs), known for their powerful zero-shot capabilities in textual tasks, show promise but struggle to naturally capture the critical structural information inherent to TAGs, limiting their direct effectiveness. To address these challenges, we propose LLM-GOOD, a general framework that effectively combines the strengths of LLMs and GNNs to enhance data efficiency in graph OOD detection. Specifically, we first leverage LLMs' strong zero-shot capabilities to filter out likely OOD nodes, significantly reducing the human annotation burden. To minimize the usage and cost of the LLM, we employ it only to annotate a small subset of unlabeled nodes. We then train a lightweight GNN filter using these noisy labels, enabling efficient predictions of ID status for all other unlabeled nodes by leveraging both textual and structural information. After obtaining node embeddings from the GNN filter, we can apply informativeness-based methods to select the most valuable nodes for precise human annotation. Finally, we train the target ID classifier using these accurately annotated ID nodes. Extensive experiments on four real-world TAG datasets demonstrate that LLM-GOOD significantly reduces human annotation costs and outperforms state-of-the-art baselines in terms of both ID classification accuracy and OOD detection performance.

LEGO-Learn: Label-Efficient Graph Open-Set Learning

How can we train graph-based models to recognize unseen classes while keeping labeling costs low? Graph open-set learning (GOL) and out-of-distribution (OOD) detection aim to address this challenge by training models that can accurately classify known, in-distribution (ID) classes while identifying and handling previously unseen classes during inference. It is critical for high-stakes, real-world applications where models frequently encounter unexpected data, including finance, security, and healthcare. However, current GOL methods assume access to many labeled ID samples, which is unrealistic for large-scale graphs due to high annotation costs. In this paper, we propose LEGO-Learn (Label-Efficient Graph Open-set Learning), a novel framework that tackles open-set node classification on graphs within a given label budget by selecting the most informative ID nodes. LEGO-Learn employs a GNN-based filter to identify and exclude potential OOD nodes and then select highly informative ID nodes for labeling using the K-Medoids algorithm. To prevent the filter from discarding valuable ID examples, we introduce a classifier that differentiates between the C known ID classes and an additional class representing OOD nodes (hence, a C+1 classifier). This classifier uses a weighted cross-entropy loss to balance the removal of OOD nodes while retaining informative ID nodes. Experimental results on four real-world datasets demonstrate that LEGO-Learn significantly outperforms leading methods, with up to a 6.62% improvement in ID classification accuracy and a 7.49% increase in AUROC for OOD detection.

Multivariate Time Series Classification with Hierarchical Variational Graph Pooling

Over the past decade, multivariate time series classification (MTSC) has received great attention with the advance of sensing techniques. Current deep learning methods for MTSC are based on convolutional and recurrent neural network, with the assumption that time series variables have the same effect to each other. Thus they cannot model the pairwise dependencies among variables explicitly. What's more, current spatial-temporal modeling methods based on GNNs are inherently flat and lack the capability of aggregating node information in a hierarchical manner. To address this limitation and attain expressive global representation of MTS, we propose a graph pooling based framework MTPool and view MTSC task as graph classification task. With graph structure learning and temporal convolution, MTS slices are converted to graphs and spatial-temporal features are extracted. Then, we propose a novel graph pooling method, which uses an ``encoder-decoder'' mechanism to generate adaptive centroids for cluster assignments. GNNs and graph pooling layers are used for joint graph representation learning and graph coarsening. With multiple graph pooling layers, the input graphs are hierachically coarsened to one node. Finally, differentiable classifier takes this coarsened one-node graph as input to get the final predicted class. Experiments on 10 benchmark datasets demonstrate MTPool outperforms state-of-the-art methods in MTSC tasks.

Modeling Complex Spatial Patterns with Temporal Features via Heterogenous Graph Embedding Networks

Multivariate time series (MTS) forecasting is an important problem in many fields. Accurate forecasting results can effectively help decision-making. Variables in MTS have rich relations among each other and the value of each variable in MTS depends both on its historical values and on other variables. These rich relations can be static and predictable or dynamic and latent. Existing methods do not incorporate these rich relational information into modeling or only model certain relation among MTS variables. To jointly model rich relations among variables and temporal dependencies within the time series, a novel end-to-end deep learning model, termed Multivariate Time Series Forecasting via Heterogenous Graph Neural Networks (MTHetGNN) is proposed in this paper. To characterize rich relations among variables, a relation embedding module is introduced in our model, where each variable is regarded as a graph node and each type of edge represents a specific relationship among variables or one specific dynamic update strategy to model the latent dependency among variables. In addition, convolutional neural network (CNN) filters with different perception scales are used for time series feature extraction, which is used to generate the feature of each node. Finally, heterogenous graph neural networks are adopted to handle the complex structural information generated by temporal embedding module and relation embedding module. Three benchmark datasets from the real world are used to evaluate the proposed MTHetGNN and the comprehensive experiments show that MTHetGNN achieves state-of-the-art results in MTS forecasting task.

Parallel Extraction of Long-term Trends and Short-term Fluctuation Framework for Multivariate Time Series Forecasting

Multivariate time series forecasting is widely used in various fields. Reasonable prediction results can assist people in planning and decision-making, generate benefits and avoid risks. Normally, there are two characteristics of time series, that is, long-term trend and short-term fluctuation. For example, stock prices will have a long-term upward trend with the market, but there may be a small decline in the short term. These two characteristics are often relatively independent of each other. However, the existing prediction methods often do not distinguish between them, which reduces the accuracy of the prediction model. In this paper, a MTS forecasting framework that can capture the long-term trends and short-term fluctuations of time series in parallel is proposed. This method uses the original time series and its first difference to characterize long-term trends and short-term fluctuations. Three prediction sub-networks are constructed to predict long-term trends, short-term fluctuations and the final value to be predicted. In the overall optimization goal, the idea of multi-task learning is used for reference, which is to make the prediction results of long-term trends and short-term fluctuations as close to the real values as possible while requiring to approximate the values to be predicted. In this way, the proposed method uses more supervision information and can more accurately capture the changing trend of the time series, thereby improving the forecasting performance.

TIMME: Twitter Ideology-detection via Multi-task Multi-relational Embedding

We aim at solving the problem of predicting people's ideology, or political tendency. We estimate it by using Twitter data, and formalize it as a classification problem. Ideology-detection has long been a challenging yet important problem. Certain groups, such as the policy makers, rely on it to make wise decisions. Back in the old days when labor-intensive survey-studies were needed to collect public opinions, analyzing ordinary citizens' political tendencies was uneasy. The rise of social medias, such as Twitter, has enabled us to gather ordinary citizen's data easily. However, the incompleteness of the labels and the features in social network datasets is tricky, not to mention the enormous data size and the heterogeneousity. The data differ dramatically from many commonly-used datasets, thus brings unique challenges. In our work, first we built our own datasets from Twitter. Next, we proposed TIMME, a multi-task multi-relational embedding model, that works efficiently on sparsely-labeled heterogeneous real-world dataset. It could also handle the incompleteness of the input features. Experimental results showed that TIMME is overall better than the state-of-the-art models for ideology detection on Twitter. Our findings include: links can lead to good classification outcomes without text; conservative voice is under-represented on Twitter; follow is the most important relation to predict ideology; retweet and mention enhance a higher chance of like, etc. Last but not least, TIMME could be extended to other datasets and tasks in theory.

Hierarchical Large-scale Graph Similarity Computation via Graph Coarsening and Matching

In this work, we focus on large graph similarity computation problem and propose a novel ``embedding-coarsening-matching'' learning framework, which outperforms state-of-the-art methods in this task and has significant improvement in time efficiency. Graph similarity computation for metrics such as Graph Edit Distance (GED) is typically NP-hard, and existing heuristics-based algorithms usually achieves a unsatisfactory trade-off between accuracy and efficiency. Recently the development of deep learning techniques provides a promising solution for this problem by a data-driven approach which trains a network to encode graphs to their own feature vectors and computes similarity based on feature vectors. These deep-learning methods can be classified to two categories, embedding models and matching models. Embedding models such as GCN-Mean and GCN-Max, which directly map graphs to respective feature vectors, run faster but the performance is usually poor due to the lack of interactions across graphs. Matching models such as GMN, whose encoding process involves interaction across the two graphs, are more accurate but interaction between whole graphs brings a significant increase in time consumption (at least quadratic time complexity over number of nodes). Inspired by large biological molecular identification where the whole molecular is first mapped to functional groups and then identified based on these functional groups, our ``embedding-coarsening-matching'' learning framework first embeds and coarsens large graphs to coarsened graphs with denser local topology and then matching mechanism is deployed on the coarsened graphs for the final similarity scores. Detailed experiments have been conducted and the results demonstrate the efficiency and effectiveness of our proposed framework.