arXiv subjects

Junjie Wu

Publications and source records attributed to Junjie Wu.

At least 73 records · Page 4Linked to original sources

arXiv2025

Graph-R1: Incentivizing the Zero-Shot Graph Learning Capability in LLMs via Explicit Reasoning

Generalizing to unseen graph tasks without task-pecific supervision remains challenging. Graph Neural Networks (GNNs) are limited by fixed label spaces, while Large Language Models (LLMs) lack structural inductive biases. Recent advances in Large Reasoning Models (LRMs) provide a zero-shot alternative via explicit, long chain-of-thought reasoning. Inspired by this, we propose a GNN-free approach that reformulates graph tasks--node classification, link prediction, and graph classification--as textual reasoning problems solved by LRMs. We introduce the first datasets with detailed reasoning traces for these tasks and develop Graph-R1, a reinforcement learning framework that leverages task-specific rethink templates to guide reasoning over linearized graphs. Experiments demonstrate that Graph-R1 outperforms state-of-the-art baselines in zero-shot settings, producing interpretable and effective predictions. Our work highlights the promise of explicit reasoning for graph learning and provides new resources for future research.

cs.LG↗

arXiv2025

Ref-Long: Benchmarking the Long-context Referencing Capability of Long-context Language Models

Long-context language models (LCLMs) have exhibited impressive capabilities in long-context understanding tasks. Among these, long-context referencing -- a crucial task that requires LCLMs to attribute items of interest to specific parts of long-context data -- remains underexplored. To bridge this gap, this paper proposes Referencing Evaluation for Long-context Language Models (Ref-Long), a novel benchmark designed to assess the long-context referencing capability of LCLMs. Specifically, Ref-Long requires LCLMs to identify the indexes of documents that reference a specific key, emphasizing contextual relationships between the key and the documents over simple retrieval. Based on the task design, we construct three subsets ranging from synthetic to realistic scenarios to form the Ref-Long benchmark. Experimental results of 13 LCLMs reveal significant shortcomings in long-context referencing, even among advanced models like GPT-4o. To further investigate these challenges, we conduct comprehensive analyses, including human evaluations, task format adjustments, fine-tuning experiments, and error analyses, leading to several key insights. Our data and code can be found in https://github. com/wujunjie1998/Ref-Long.

cs.CL↗

arXiv2025

Learning-based Privacy-Preserving Graph Publishing Against Sensitive Link Inference Attacks

Publishing graph data is widely desired to enable a variety of structural analyses and downstream tasks. However, it also potentially poses severe privacy leakage, as attackers may leverage the released graph data to launch attacks and precisely infer private information such as the existence of hidden sensitive links in the graph. Prior studies on privacy-preserving graph data publishing relied on heuristic graph modification strategies and it is difficult to determine the graph with the optimal privacy--utility trade-off for publishing. In contrast, we propose the first privacy-preserving graph structure learning framework against sensitive link inference attacks, named PPGSL, which can automatically learn a graph with the optimal privacy--utility trade-off. The PPGSL operates by first simulating a powerful surrogate attacker conducting sensitive link attacks on a given graph. It then trains a parameterized graph to defend against the simulated adversarial attacks while maintaining the favorable utility of the original graph. To learn the parameters of both parts of the PPGSL, we introduce a secure iterative training protocol. It can enhance privacy preservation and ensure stable convergence during the training process, as supported by the theoretical proof. Additionally, we incorporate multiple acceleration techniques to improve the efficiency of the PPGSL in handling large-scale graphs. The experimental results confirm that the PPGSL achieves state-of-the-art privacy--utility trade-off performance and effectively thwarts various sensitive link inference attacks.

cs.CR↗

arXiv2025

Unified Triplet-Level Hallucination Evaluation for Large Vision-Language Models

Despite the outstanding performance in vision-language reasoning, Large Vision-Language Models (LVLMs) might generate hallucinated contents that do not exist in the given image. Most existing LVLM hallucination benchmarks are constrained to evaluate the object-related hallucinations. However, the potential hallucination on the relations between two objects, i.e., relation hallucination, still lacks investigation. To remedy that, we design a unified framework to measure the object and relation hallucination in LVLMs simultaneously. The core idea of our framework is to evaluate hallucinations via (object, relation, object) triplets extracted from LVLMs' responses, making it easily generalizable to different vision-language tasks. Based on our framework, we further introduce Tri-HE, a novel Triplet-level Hallucination Evaluation benchmark which can be used to study both object and relation hallucination at the same time. With comprehensive evaluations on Tri-HE, we observe that the relation hallucination issue is even more serious than object hallucination among existing LVLMs, highlighting a previously neglected problem towards reliable LVLMs. Moreover, based on our findings, we design a simple training-free approach that effectively mitigates hallucinations for LVLMs. Our dataset and code for the reproduction of our experiments are available publicly at https://github.com/wujunjie1998/Tri-HE.

cs.CV↗

arXiv2025

SimDC: A High-Fidelity Device Simulation Platform for Device-Cloud Collaborative Computing

The advent of edge intelligence and escalating concerns for data privacy protection have sparked a surge of interest in device-cloud collaborative computing. Large-scale device deployments to validate prototype solutions are often prohibitively expensive and practically challenging, resulting in a pronounced demand for simulation tools that can emulate realworld scenarios. However, existing simulators predominantly rely solely on high-performance servers to emulate edge computing devices, overlooking (1) the discrepancies between virtual computing units and actual heterogeneous computing devices and (2) the simulation of device behaviors in real-world environments. In this paper, we propose a high-fidelity device simulation platform, called SimDC, which uses a hybrid heterogeneous resource and integrates high-performance servers and physical mobile phones. Utilizing this platform, developers can simulate numerous devices for functional testing cost-effectively and capture precise operational responses from varied real devices. To simulate real behaviors of heterogeneous devices, we offer a configurable device behavior traffic controller that dispatches results on devices to the cloud using a user-defined operation strategy. Comprehensive experiments on the public dataset show the effectiveness of our simulation platform and its great potential for application. The code is available at https://github.com/opas-lab/olearning-sim.

cs.DC↗

arXiv2025

DALIP: Distribution Alignment-based Language-Image Pre-Training for Domain-Specific Data

Recently, Contrastive Language-Image Pre-training (CLIP) has shown promising performance in domain-specific data (e.g., biology), and has attracted increasing research attention. Existing works generally focus on collecting extensive domain-specific data and directly tuning the original CLIP models. Intuitively, such a paradigm takes no full consideration of the characteristics lying in domain-specific data (e.g., fine-grained nature of biological data) and so limits model capability, while mostly losing the original ability of CLIP in the general domain. In this paper, we propose a Distribution Alignment-based Language-Image Pre-Training (DALIP) method for biological data. Specifically, DALIP optimizes CLIP models by matching the similarity between feature distribution of image-text pairs instead of the original [cls] token, which can capture rich yet effective information inherent in image-text pairs as powerful representations, and so better cope with fine-grained nature of biological data. Particularly, our DALIP efficiently approximates feature distribution via its first- and second-order statistics, while presenting a Multi-head Brownian Distance Covariance (MBDC) module to acquire second-order statistics of token features efficiently. Furthermore, we collect a new dataset for plant domain (e.g., specific data in biological domain) comprising 10M plant data with 3M general-domain data (namely PlantMix-13M) according to data mixing laws. Extensive experiments show that DALIP clearly outperforms existing CLIP counterparts in biological domain, while well generalizing to remote sensing and medical imaging domains. Besides, our PlantMix-13M dataset further boosts performance of DALIP in plant domain, while preserving model ability in general domain.

cs.CV↗

arXiv2025

BeMERC: Behavior-Aware MLLM-based Framework for Multimodal Emotion Recognition in Conversation

Multimodal emotion recognition in conversation (MERC), the task of identifying the emotion label for each utterance in a conversation, is vital for developing empathetic machines. Current MLLM-based MERC studies focus mainly on capturing the speaker's textual or vocal characteristics, but ignore the significance of video-derived behavior information. Different from text and audio inputs, learning videos with rich facial expression, body language and posture, provides emotion trigger signals to the models for more accurate emotion predictions. In this paper, we propose a novel behavior-aware MLLM-based framework (BeMERC) to incorporate speaker's behaviors, including subtle facial micro-expression, body language and posture, into a vanilla MLLM-based MERC model, thereby facilitating the modeling of emotional dynamics during a conversation. Furthermore, BeMERC adopts a two-stage instruction tuning strategy to extend the model to the conversations scenario for end-to-end training of a MERC predictor. Experiments demonstrate that BeMERC achieves superior performance than the state-of-the-art methods on two benchmark datasets, and also provides a detailed discussion on the significance of video-derived behavior information in MERC.

cs.CL↗

arXiv2025

LaERC-S: Improving LLM-based Emotion Recognition in Conversation with Speaker Characteristics

Emotion recognition in conversation (ERC), the task of discerning human emotions for each utterance within a conversation, has garnered significant attention in human-computer interaction systems. Previous ERC studies focus on speaker-specific information that predominantly stems from relationships among utterances, which lacks sufficient information around conversations. Recent research in ERC has sought to exploit pre-trained large language models (LLMs) with speaker modelling to comprehend emotional states. Although these methods have achieved encouraging results, the extracted speaker-specific information struggles to indicate emotional dynamics. In this paper, motivated by the fact that speaker characteristics play a crucial role and LLMs have rich world knowledge, we present LaERC-S, a novel framework that stimulates LLMs to explore speaker characteristics involving the mental state and behavior of interlocutors, for accurate emotion predictions. To endow LLMs with this knowledge information, we adopt the two-stage learning to make the models reason speaker characteristics and track the emotion of the speaker in complex conversation scenarios. Extensive experiments on three benchmark datasets demonstrate the superiority of LaERC-S, reaching the new state-of-the-art.

cs.CL↗

arXiv2025

Understanding LLMs' Fluid Intelligence Deficiency: An Analysis of the ARC Task

While LLMs have exhibited strong performance on various NLP tasks, it is noteworthy that most of these tasks rely on utilizing the vast amount of knowledge encoded in LLMs' parameters, rather than solving new problems without prior knowledge. In cognitive research, the latter ability is referred to as fluid intelligence, which is considered to be critical for assessing human intelligence. Recent research on fluid intelligence assessments has highlighted significant deficiencies in LLMs' abilities. In this paper, we analyze the challenges LLMs face in demonstrating fluid intelligence through controlled experiments, using the most representative ARC task as an example. Our study revealed three major limitations in existing LLMs: limited ability for skill composition, unfamiliarity with abstract input formats, and the intrinsic deficiency of left-to-right decoding. Our data and code can be found in https://wujunjie1998.github.io/araoc-benchmark.github.io/.

cs.AI↗

arXiv2025

Error-mitigated entanglement-assisted quantum process tomography

In the era of noisy intermediate-scale quantum computing, it is of crucial importance to verify quantum processes and extract information. Quantum process tomography is a typical approach, however, both resource-intensive and vulnerable to state preparation and measurement errors. Here, we propose an error-mitigated entanglement-assisted quantum process tomography (EM-EAPT) framework to address these limitations. By leveraging a maximally entangled state to reduce state preparation complexity and integrating error mitigation techniques, our method significantly enhances robustness against SPAM errors. Experimental validation on a superconducting processor demonstrates the efficacy of EM-EAPT for two-qubit and three-qubit quantum processes. Results show more accurate average gate fidelities close to the realistic estimation, achieving 98.1$\pm$ 0.03% for a CNOT gate and 88.1%$\pm$ 0.04% for a cascaded CNOT process after error mitigation, compared to non-mitigated implementations. This work advances practical quantum verification tools for NISQ devices, enabling higher-fidelity characterization of quantum processes under realistic noise conditions.

quant-ph↗

arXiv2025

The Stochastic Parrot on LLM's Shoulder: A Summative Assessment of Physical Concept Understanding

In a systematic way, we investigate a widely asked question: Do LLMs really understand what they say?, which relates to the more familiar term Stochastic Parrot. To this end, we propose a summative assessment over a carefully designed physical concept understanding task, PhysiCo. Our task alleviates the memorization issue via the usage of grid-format inputs that abstractly describe physical phenomena. The grids represents varying levels of understanding, from the core phenomenon, application examples to analogies to other abstract patterns in the grid world. A comprehensive study on our task demonstrates: (1) state-of-the-art LLMs, including GPT-4o, o1 and Gemini 2.0 flash thinking, lag behind humans by ~40%; (2) the stochastic parrot phenomenon is present in LLMs, as they fail on our grid task but can describe and recognize the same concepts well in natural language; (3) our task challenges the LLMs due to intrinsic difficulties rather than the unfamiliar grid format, as in-context learning and fine-tuning on same formatted data added little to their performance.

cs.CL↗

arXiv2025

E2LVLM:Evidence-Enhanced Large Vision-Language Model for Multimodal Out-of-Context Misinformation Detection

Recent studies in Large Vision-Language Models (LVLMs) have demonstrated impressive advancements in multimodal Out-of-Context (OOC) misinformation detection, discerning whether an authentic image is wrongly used in a claim. Despite their success, the textual evidence of authentic images retrieved from the inverse search is directly transmitted to LVLMs, leading to inaccurate or false information in the decision-making phase. To this end, we present E2LVLM, a novel evidence-enhanced large vision-language model by adapting textual evidence in two levels. First, motivated by the fact that textual evidence provided by external tools struggles to align with LVLMs inputs, we devise a reranking and rewriting strategy for generating coherent and contextually attuned content, thereby driving the aligned and effective behavior of LVLMs pertinent to authentic images. Second, to address the scarcity of news domain datasets with both judgment and explanation, we generate a novel OOC multimodal instruction-following dataset by prompting LVLMs with informative content to acquire plausible explanations. Further, we develop a multimodal instruction-tuning strategy with convincing explanations for beyond detection. This scheme contributes to E2LVLM for multimodal OOC misinformation detection and explanation. A multitude of experiments demonstrate that E2LVLM achieves superior performance than state-of-the-art methods, and also provides compelling rationales for judgments.

cs.LG↗

arXiv2025

Bridging Traffic State and Trajectory for Dynamic Road Network and Trajectory Representation Learning

Effective urban traffic management is vital for sustainable city development, relying on intelligent systems with machine learning tasks such as traffic flow prediction and travel time estimation. Traditional approaches usually focus on static road network and trajectory representation learning, and overlook the dynamic nature of traffic states and trajectories, which is crucial for downstream tasks. To address this gap, we propose TRACK, a novel framework to bridge traffic state and trajectory data for dynamic road network and trajectory representation learning. TRACK leverages graph attention networks (GAT) to encode static and spatial road segment features, and introduces a transformer-based model for trajectory representation learning. By incorporating transition probabilities from trajectory data into GAT attention weights, TRACK captures dynamic spatial features of road segments. Meanwhile, TRACK designs a traffic transformer encoder to capture the spatial-temporal dynamics of road segments from traffic state data. To further enhance dynamic representations, TRACK proposes a co-attentional transformer encoder and a trajectory-traffic state matching task. Extensive experiments on real-life urban traffic datasets demonstrate the superiority of TRACK over state-of-the-art baselines. Case studies confirm TRACK's ability to capture spatial-temporal dynamics effectively.

cs.LG↗

arXiv2025

Learning Universal Multi-level Market Irrationality Factors to Improve Stock Return Forecasting

Recent years have witnessed the perfect encounter of deep learning and quantitative trading has achieved great success in stock investment. Numerous deep learning-based models have been developed for forecasting stock returns, leveraging the powerful representation capabilities of neural networks to identify patterns and factors influencing stock prices. These models can effectively capture general patterns in the market, such as stock price trends, volume-price relationships, and time variations. However, the impact of special irrationality factors -- such as market sentiment, speculative behavior, market manipulation, and psychological biases -- have not been fully considered in existing deep stock forecasting models due to their relative abstraction as well as lack of explicit labels and data description. To fill this gap, we propose UMI, a Universal multi-level Market Irrationality factor model to enhance stock return forecasting. The UMI model learns factors that can reflect irrational behaviors in market from both individual stock and overall market levels. For the stock-level, UMI construct an estimated rational price for each stock, which is cointegrated with the stock's actual price. The discrepancy between the actual and the rational prices serves as a factor to indicate stock-level irrational events. Additionally, we define market-level irrational behaviors as anomalous synchronous fluctuations of stocks within a market. Using two self-supervised representation learning tasks, i.e., sub-market comparative learning and market synchronism prediction, the UMI model incorporates market-level irrationalities into a market representation vector, which is then used as the market-level irrationality factor.

cs.LG↗

arXiv2025

Exclusion of a direct progenitor detection for the Type Ic SN 2017ein based on late-time observations

To date, SN 2017ein is the only Type Ic supernova with a directly identified progenitor candidate. This candidate points to a very massive ($>$45 $M_\odot$) Wolf-Rayet progenitor, but its disappearance after the explosion of SN 2017ein remains unconfirmed. In this work, we revisit SN 2017ein in late-time images acquired by the Hubble Space Telescope (HST) at 2.4--3.8 yrs after peak brightness. We find this source has not disappeared and its brightness and color remain almost the same as in the pre-explosion images. Thus, we conclude that the pre-explosion source is not the genuine progenitor of SN 2017ein. We exclude the possibility that it is a companion star of the progenitor, since it has a much lower extinction than SN 2017ein; its color is also inconsistent with a star cluster, indicated by the newly added magnitude limit in F336W, apart from F555W and F814W. We suggest, therefore, this source is an unrelated star in chance alignment with SN 2017ein. Based on the low ejecta mass, we propose that SN 2017ein is most likely originated from a moderately massive star with $M_{\rm ini}$ $\sim$ 8--20 $M_\odot$, stripped by binary interaction, rather than a very massive Wolf-Rayet progenitor.

astro-ph.HE↗

arXiv2024

LLMs as Zero-shot Graph Learners: Alignment of GNN Representations with LLM Token Embeddings

Zero-shot graph machine learning, especially with graph neural networks (GNNs), has garnered significant interest due to the challenge of scarce labeled data. While methods like self-supervised learning and graph prompt learning have been extensively explored, they often rely on fine-tuning with task-specific labels, limiting their effectiveness in zero-shot scenarios. Inspired by the zero-shot capabilities of instruction-fine-tuned large language models (LLMs), we introduce a novel framework named Token Embedding-Aligned Graph Language Model (TEA-GLM) that leverages LLMs as cross-dataset and cross-task zero-shot learners for graph machine learning. Concretely, we pretrain a GNN, aligning its representations with token embeddings of an LLM. We then train a linear projector that transforms the GNN's representations into a fixed number of graph token embeddings without tuning the LLM. A unified instruction is designed for various graph tasks at different levels, such as node classification (node-level) and link prediction (edge-level). These design choices collectively enhance our method's effectiveness in zero-shot learning, setting it apart from existing methods. Experiments show that our graph token embeddings help the LLM predictor achieve state-of-the-art performance on unseen datasets and tasks compared to other methods using LLMs as predictors.

cs.LG↗

arXiv2024

GaussianAD: Gaussian-Centric End-to-End Autonomous Driving

Vision-based autonomous driving shows great potential due to its satisfactory performance and low costs. Most existing methods adopt dense representations (e.g., bird's eye view) or sparse representations (e.g., instance boxes) for decision-making, which suffer from the trade-off between comprehensiveness and efficiency. This paper explores a Gaussian-centric end-to-end autonomous driving (GaussianAD) framework and exploits 3D semantic Gaussians to extensively yet sparsely describe the scene. We initialize the scene with uniform 3D Gaussians and use surrounding-view images to progressively refine them to obtain the 3D Gaussian scene representation. We then use sparse convolutions to efficiently perform 3D perception (e.g., 3D detection, semantic map construction). We predict 3D flows for the Gaussians with dynamic semantics and plan the ego trajectory accordingly with an objective of future scene forecasting. Our GaussianAD can be trained in an end-to-end manner with optional perception labels when available. Extensive experiments on the widely used nuScenes dataset verify the effectiveness of our end-to-end GaussianAD on various tasks including motion planning, 3D occupancy prediction, and 4D occupancy forecasting. Code: https://github.com/wzzheng/GaussianAD.

cs.CV↗

arXiv2024

Constraining the progenitor of the nearby Type II-P SN 2024ggi with environmental analysis

The progenitors of Type II-P supernovae (SN) have been confirmed to be red supergiants. However, the upper mass limit of the directly probed progenitors is much lower than that predicted by current theories, and the accurate determination of the progenitor masses is key to understand the final fate of massive stars. Located at a distance of only 6.72 Mpc, the Type II-P SN 2024ggi is one of the closest SN in the last decade. Previous studies have analyzed its progenitor by direct detection, but the derived progenitor mass may be influenced by the very uncertain circumstellar extinction and pulsational brightness variability. In this work, we try to constrain the progenitor mass with an environmental analysis based on images from the Hubble Space Telescope. We found that stars in the progenitor environment have a uniform spatial distribution without significant clumpiness, and we derived the star formation history of the environment with a hierarchical Bayesian method. The progenitor is associated with the youngest population in the SN environment with an age of log($t$/yr) = 7.41 (i.e. 25.7 Myr), which corresponds to an initial mass of $10.2^{+0.06}_{-0.09}$ $M_\odot$. Our work provides an independent measurement of the progenitor mass, which is not affected by circumstellar extinction and pulsational brightness variability.

astro-ph.SR↗