arXiv subjects

Jingyu Wu

Publications and source records attributed to Jingyu Wu.

18 recordsLinked to original sources

arXiv2026

CAGE: Coherence-Aware Graph Encoding for Retrieval-Augmented Generation

Traditional Retrieval-Augmented Generation (RAG) systems score each passage independently against the query, assembling context sets that may be individually relevant yet collectively incoherent. We introduce Coherence-Aware Graph Encoding (CAGE), a reranking framework that models "between-chunk coherence" across four dimensions: Intra-Domain Relevance, Noise Resistance, Informational Bonding, and Factual Consistency. Our pipeline transforms retrieved passages into directed heterogeneous entity graphs, amplifies factual anchors via min-out-degree reweighting, encodes structural patterns through a Relational Graph Convolutional Network, and fuses inter-chunk coherence with query relevance for final ranking. Evaluated across four multi-hop benchmarks, CAGE matches or outperforms strong baselines including monoT5 in Recall@5 on bridge-dominated datasets and consistently improves downstream Exact Match, demonstrating that structurally coherent context yields more precise answers even when retrieval recall is comparable or lower.

cs.CL↗

arXiv2026

Towards Valid B-Rep Generation: Training-Free Wireframe Anomaly Detection and Repair

Multi-stage boundary representation (B-Rep) generation leverages intermediate wireframes to synthesize CAD models. However, geometric and topological risks in these wireframes -- such as self-intersections, edge collapses, and disconnected vertices -- can propagate to invalid final B-Reps. Mitigating such failures by retraining large generative models is computationally prohibitive. We propose Wireframe Detection and Repair (WDR), a training-free framework that intervenes at the intermediate wireframe stage to improve downstream B-Rep validity. WDR features a Geometric-Topology Anomaly Detector (GTAD) that combines parallel VLM-based coarse screening with geometric and topological detectors to predict downstream invalidity risk and route generation to dedicated branches. An Energy-Guided Geometric-Topology Repair (EGGTR) module then performs detector-triggered guided regeneration through geometry and topology branches. By scaling test-time computation via Energy-Guided Resampling and training-free guidance for diffusion models, WDR can be integrated into autoregressive and diffusion pipelines without retraining. Extensive experiments demonstrate consistent improvements in kernel-checked validity while largely retaining the measured diversity and distributional quality of synthesized CAD models. The code will be made publicly available upon acceptance.

cs.CV↗

arXiv2026

$L^{2}$-Supercritical Nonlinear Klein-Gordon System with Quadratic Asymmetric Interaction

In this paper we investigate the global existence, blow-up and standing waves for the $L^{2}$-supercritical nonlinear Klein-Gordon equations with quadratic asymmetric interaction (LSNKG). First, by introducing a suitable auxiliary functional, using concavity analysis and virial estimates, we obtain a finite time blow-up result for solutions to the Cauchy problem of (LSNKG) when the initial energy is negative. Next, by defining appropriate functionals, manifolds and a constrained variational problem, we employ variational method and the Lagrange multiplier method to derive the existence of ground state solutions for the corresponding nonlinear elliptic (steady-state) system, thereby to establish the existence of standing wave with the ground state for (LSNKG). Then, using the variational characterization of the ground state solutions and constructing invariant sets under the flow generated by the Cauchy problem for (LSNKG), we combine the potential well argument with concavity analysis to establish a sharp threshold between blow-up in finite time and global existence. Finally, by exploiting the variational characterization of the ground state, introducing appropriate scalings, and choosing suitable initial data based on the ground state, we justify the instability of standing wave with the ground state for (LSNKG).

math.AP↗

arXiv2026

SAFARI: Scaling Long Horizon Agentic Fault Attribution via Active Investigation

As autonomous agents tackle increasingly complex multi-step, multi-agent tasks, their execution trajectories have scaled beyond the constraints of even the largest context windows. Current methods for effectively diagnosing agent failures load the full trajectory into an LLM's context window, which suffers from attention dilution and fails when agentic traces inevitably exceed context limits. To address this, we introduce SAFARI (Scaling long-horizon Agentic Fault AttRibution via active Investigation), a framework that replaces linear context loading with a tool-augmented diagnostic loop. By equipping LLMs with a specialized toolbox to read and search trajectory segments alongside a persistent Short-Term Memory (STM) for cross-turn reasoning, SAFARI effectively decouples diagnostic accuracy from architectural context limits. Our experiments demonstrate that SAFARI outperforms state-of-the-art results by 20% on the Who&When dataset within a 1M token budget, and by 19% on TRAIL GAIA subset on a 25K token budget. Most significantly, SAFARI maintains a 0.58 precision even when the target fault resides 5x beyond the model's native context window, a scenario where traditional evaluators fail entirely.

cs.AI↗

arXiv2026

Quantum Property Testing for Bounded-Degree Directed Graphs

We study quantum property testing for directed graphs with maximum in-degree and out-degree bounded by some universal constant $d$. For a proximity parameter $\varepsilon$, we show that any property that can be tested with $O_{\varepsilon,d}(1)$ queries in the classical bidirectional model, where both incoming and outgoing edges are accessible, can also be tested in the quantum unidirectional model, where only outgoing edges are accessible, using $n^{1/2 - \Omega_{\varepsilon,d}(1)}$ queries. This yields an almost quadratic quantum speedup over the best known classical algorithms in the unidirectional model. Moreover, we prove that our transformation is almost tight by giving an explicit property $P_\varepsilon$ that is $\varepsilon$-testable within $O_\varepsilon(1)$ classical queries in the bidirectional model, but requires $\widetilde{\Omega}(n^{1/2-f'(\varepsilon)})$ quantum queries in the unidirectional model, where $f'(\varepsilon)$ is a function that approaches $0$ as $\varepsilon$ approaches $0$. As a byproduct, we show that in the unidirectional model, the number of occurrences of any constant-size subgraph $H$ can be approximated up to additive error $\delta n$ using $o(\sqrt{n})$ quantum queries.

quant-ph↗

arXiv2026

Bridging Time and Space: Decoupled Spatio-Temporal Alignment for Video Grounding

Spatio-Temporal Video Grounding requires jointly localizing target objects across both temporal and spatial dimensions based on natural language queries, posing fundamental challenges for existing Multimodal Large Language Models (MLLMs). We identify two core challenges: \textit{entangled spatio-temporal alignment}, arising from coupling two heterogeneous sub-tasks within the same autoregressive output space, and \textit{dual-domain visual token redundancy}, where target objects exhibit simultaneous temporal and spatial sparsity, rendering the overwhelming majority of visual tokens irrelevant to the grounding query. To address these, we propose \textbf{Bridge-STG}, an end-to-end framework that decouples temporal and spatial localization while maintaining semantic coherence. While decoupling is the natural solution to this entanglement, it risks creating a semantic gap between the temporal MLLM and the spatial decoder. Bridge-STG resolves this through two pivotal designs: the \textbf{Spatio-Temporal Semantic Bridging (STSB)} mechanism with Explicit Temporal Alignment (ETA) distills the MLLM's temporal reasoning context into enriched bridging queries as a robust semantic interface; and the \textbf{Query-Guided Spatial Localization (QGSL)} module leverages these queries to drive a purpose-built spatial decoder with multi-layer interactive queries and positive/negative frame sampling, jointly eliminating dual-domain visual token redundancy. Extensive experiments across multiple benchmarks demonstrate that Bridge-STG achieves state-of-the-art performance among MLLM-based methods. Bridge-STG improves average m\_vIoU from $26.4$ to $34.3$ on VidSTG and demonstrates strong cross-task transfer across various fine-grained video understanding tasks under a unified multi-task training regime.

cs.CV↗

arXiv2026

DF-RAG: Query-Aware Diversity for Retrieval-Augmented Generation

Retrieval-augmented generation (RAG) is a common technique for grounding language model outputs in domain-specific information. However, RAG is often challenged by reasoning-intensive question-answering (QA), since common retrieval methods like cosine similarity maximize relevance at the cost of introducing redundant content, which can reduce information recall. To address this, we introduce Diversity-Focused Retrieval-Augmented Generation (DF-RAG), which systematically incorporates diversity into the retrieval step to improve performance on complex, reasoning-intensive QA benchmarks. DF-RAG builds upon the Maximal Marginal Relevance framework to select information chunks that are both relevant to the query and maximally dissimilar from each other. A key innovation of DF-RAG is its ability to optimize the level of diversity for each query dynamically at test time without requiring any additional fine-tuning or prior information. We show that DF-RAG improves F1 performance on reasoning-intensive QA benchmarks by 4-10 percent over vanilla RAG using cosine similarity and also outperforms other established baselines. Furthermore, we estimate an Oracle ceiling of up to 18 percent absolute F1 gains over vanilla RAG, of which DF-RAG captures up to 91.3 percent.

cs.CL↗

arXiv2025

OmniGround: A Comprehensive Spatio-Temporal Grounding Benchmark for Real-World Complex Scenarios

Spatio-Temporal Video Grounding (STVG) aims to localize target objects in videos based on natural language descriptions. Despite recent advances in Multimodal Large Language Models, a significant gap remains between current models and real-world demands involving diverse objects and complex queries. We attribute this to limited benchmark scope, causing models to exhibit category bias, oversimplified reasoning, and poor linguistic robustness. To address these limitations, we introduce OmniGround, a comprehensive benchmark with 3,475 videos spanning 81 categories and complex real-world queries. We propose the Forward-Backward-Refinement annotation pipeline that combines multi-directional tracking with intelligent error correction for high-quality labels. We further introduce DeepSTG, a systematic evaluation framework quantifying dataset quality across four complementary dimensions beyond superficial statistics. Evaluations reveal performance average drop of 10.4% on complex real-world scenes, particularly with small/occluded objects and intricate spatial relations. Motivated by these, we propose PG-TAF, a training-free two-stage framework decomposing STVG into high-level temporal grounding and fine-grained spatio-temporal propagation. Experiments demonstrate PG-TAF achieves 25.6% and 35.6% improvements in m\_tIoU and m\_vIoU on OmniGround with consistent gains across four benchmarks.

cs.CV↗

arXiv2025

LLM Optimization Unlocks Real-Time Pairwise Reranking

Efficiently reranking documents retrieved from information retrieval (IR) pipelines to enhance overall quality of Retrieval-Augmented Generation (RAG) system remains an important yet challenging problem. Recent studies have highlighted the importance of Large Language Models (LLMs) in reranking tasks. In particular, Pairwise Reranking Prompting (PRP) has emerged as a promising plug-and-play approach due to its usability and effectiveness. However, the inherent complexity of the algorithm, coupled with the high computational demands and latency incurred due to LLMs, raises concerns about its feasibility in real-time applications. To address these challenges, this paper presents a focused study on pairwise reranking, demonstrating that carefully applied optimization methods can significantly mitigate these issues. By implementing these methods, we achieve a remarkable latency reduction of up to 166 times, from 61.36 seconds to 0.37 seconds per query, with an insignificant drop in performance measured by Recall@k. Our study highlights the importance of design choices that were previously overlooked, such as using smaller models, limiting the reranked set, using lower precision, reducing positional bias with one-directional order inference, and restricting output tokens. These optimizations make LLM-based reranking substantially more efficient and feasible for latency-sensitive, real-world deployments.

cs.CL↗

arXiv2025

Detection of Quasi-periodic Oscillations in the $\gamma$-Ray Light Curve of 4FGL J0309.9-6058

In this work, we report, for the first time, a quasi-periodic oscillation (QPO) in the $\gamma$-ray band of 4FGL J0309.9-6058, also known as PKS 0308-611. We employed three analytical methods (the Lomb-Scargle periodogram, REDFIT, and the weighted wavelet Z-transform) to analyze the QPO signal using \textit{Fermi} $\gamma$-ray light curve data. The analysis reveals a potential QPO during MJD 57983$-$60503, with a period of approximately 550 days and a maximum local significance of 3.72$\sigma$ and global significance of 2.72$\sigma$ derived from the WWZ analysis. To validate this result, we applied Gaussian Process (GP) to the same light curve, which independently confirms the presence of QPO signal consistent with our Fourier-based results. We further extended the analysis to the full duration of the \textit{Fermi} observations, and the results consistently support and strengthen the presence of this QPO signal. Additionally, a time lag between the optical and $\gamma$-ray bands indicates separate emission regions for these two bands. Given the year-like timescale of the QPO signal and the fact that a QPO signal with local significance over 3$\sigma$ for full \textit{Fermi}-LAT observed time, we suggest that the QPO is most likely caused by a precessing jet.

astro-ph.HE↗

arXiv2025

Harmonizing Diverse Models: A Layer-wise Merging Strategy for Consistent Generation

Retrieval-Augmented Generation (RAG) systems leverage Large Language Models (LLMs) to generate accurate and reliable responses that are grounded in retrieved context. However, LLMs often generate inconsistent outputs for semantically equivalent inputs, a problem compounded by the scarcity of consistency-focused training data and the limitations of current fine-tuning techniques in enhancing output consistency. We propose a new approach combining systematic synthetic data generation, triplet loss for better embeddings, and a novel layer-wise model merging approach. Using consistency-aware weights derived from intermediate layer activations, our method effectively integrates knowledge from specialized models. Experimental results how that our merged model significantly enhances output consistency, achieving a ~47.5\% improvement in response similarity over the baseline, thus offering a practical solution for increasing the reliability of an industrial RAG system.

cs.CL↗

arXiv2025

RAFFLES: Reasoning-based Attribution of Faults for LLM Systems

The advent of complex, interconnected long-horizon LLM systems has made it incredibly tricky to identify where and when these systems break down. Evaluation capabilities that currently exist today are limited in that they often focus on simple metrics, end-to-end outcomes, and are dependent on the perspectives of humans. In order to match the increasing complexity of these many component systems, evaluation frameworks must also be able to reason, probe, iterate, and understand the nuanced logic passing through these systems. In this paper, we present RAFFLES, an offline evaluation architecture that incorporates iterative reasoning. Specifically, RAFFLES operates as an iterative, multi-component pipeline, using a central Judge to systematically identify faults and a set of specialized Evaluators to assess the quality of the candidate faults as well as rationales of the Judge. We evaluated RAFFLES with several benchmarks - the Who&When dataset to identify step-level faults in multi-agent systems and the ReasonEval datasets to diagnose step-level mathematical reasoning errors. RAFFLES outperforms strong baselines, achieving an accuracy of over 20% and 50% on the Who&When Hand-Crafted and Algorithmically-Generated datasets, and over 80% on the ReasonEval datasets. These results demonstrate a key step towards introducing automated fault detection for autonomous systems over labor-intensive manual review.

cs.AI↗

arXiv2024

Revisiting the flaring activity in early 2015 of BL Lacertae object S5 0716+714

In this work, we analyzed multi-wavelength data of the BL Lac object S5 0716+714 to investigate its emission mechanisms during a flaring state observed in early 2015. We examined the temporal behavior and broadband spectral energy distributions (SEDs) during the flare. The size of the $\gamma$-ray emission region was estimated based on the variability timescale. To explore the multi-wavelength properties of S5 0716+714, we employed three one-zone models: the SSC model, the SSC plus EC model, and the SSC plus pp interactions model, to reproduce the SEDs. Our findings indicate that while the SSC model can describe the SEDs, it requires an extreme Doppler factor. In contrast, the SSC plus EC model successfully fits the SEDs under the assumption of weak external photon fields but requires a high Doppler factor. Additionally, the SSC plus interactions model also reproduces the SEDs, with $\gamma$-ray emission originating from $\pi^{0}$ decay. However, this model leads to a jet power that exceeds the Eddington luminosity, which remains plausible due to the flaring state or the presence of a highly collimated jet.

astro-ph.HE↗

arXiv2024

The Study of Jet Formation Mechanism in Fermi Blazars

The origin of jet launching mainly comes from two mechanisms: the BZ mechanism and the BP mechanism. However, it is in debate which one is dominating in blazars. In this work, we used a sample of 937 Fermi blazars to study the jet formation mechanism. We studied the correlation between the jet power and the accretion rate, as well as the comparison between jet power estimated by spectral energy distribution (SED) fitting and that estimated by theoretical formula and radio flux density. Our results suggest that there is no correlation between jet power estimated by SED fitting and the accretion rate for BL Lacs, while a positive and weak correlation exists for flat spectrum radio quasars (FSRQs). Meanwhile, to confirm whether the BP and BZ mechanism is sufficient to launch the jet for FSRQs and BL Lacs, we compare the theoretical jet power with that estimated by SED fitting, as well as that by radio emission. We found that the jet power for most of the two subclasses estimated by SED fitting cannot be explained by either the BP or BZ mechanism. While the jet power for most FSRQs estimated by radio flux density can be explained by the BP mechanism, and most BL Lacs can be explained by the BZ mechanism. We also found that FSRQs have higher accretion rates than BL Lacs, implying different accretion disks around their central black holes: FSRQs typically have standard disks, while BL Lacs usually have advection-dominated accretion flow disks.

astro-ph.HE↗

arXiv2024

AgileIR: Memory-Efficient Group Shifted Windows Attention for Agile Image Restoration

Image Transformers show a magnificent success in Image Restoration tasks. Nevertheless, most of transformer-based models are strictly bounded by exorbitant memory occupancy. Our goal is to reduce the memory consumption of Swin Transformer and at the same time speed up the model during training process. Thus, we introduce AgileIR, group shifted attention mechanism along with window attention, which sparsely simplifies the model in architecture. We propose Group Shifted Window Attention (GSWA) to decompose Shift Window Multi-head Self Attention (SW-MSA) and Window Multi-head Self Attention (W-MSA) into groups across their attention heads, contributing to shrinking memory usage in back propagation. In addition to that, we keep shifted window masking and its shifted learnable biases during training, in order to induce the model interacting across windows within the channel. We also re-allocate projection parameters to accelerate attention matrix calculation, which we found a negligible decrease in performance. As a result of experiment, compared with our baseline SwinIR and other efficient quantization models, AgileIR keeps the performance still at 32.20 dB on Set5 evaluation dataset, exceeding other methods with tailor-made efficient methods and saves over 50% memory while a large batch size is employed.

cs.CV↗

arXiv2024

Hallu-PI: Evaluating Hallucination in Multi-modal Large Language Models within Perturbed Inputs

Multi-modal Large Language Models (MLLMs) have demonstrated remarkable performance on various visual-language understanding and generation tasks. However, MLLMs occasionally generate content inconsistent with the given images, which is known as "hallucination". Prior works primarily center on evaluating hallucination using standard, unperturbed benchmarks, which overlook the prevalent occurrence of perturbed inputs in real-world scenarios-such as image cropping or blurring-that are critical for a comprehensive assessment of MLLMs' hallucination. In this paper, to bridge this gap, we propose Hallu-PI, the first benchmark designed to evaluate Hallucination in MLLMs within Perturbed Inputs. Specifically, Hallu-PI consists of seven perturbed scenarios, containing 1,260 perturbed images from 11 object types. Each image is accompanied by detailed annotations, which include fine-grained hallucination types, such as existence, attribute, and relation. We equip these annotations with a rich set of questions, making Hallu-PI suitable for both discriminative and generative tasks. Extensive experiments on 12 mainstream MLLMs, such as GPT-4V and Gemini-Pro Vision, demonstrate that these models exhibit significant hallucinations on Hallu-PI, which is not observed in unperturbed scenarios. Furthermore, our research reveals a severe bias in MLLMs' ability to handle different types of hallucinations. We also design two baselines specifically for perturbed scenarios, namely Perturbed-Reminder and Perturbed-ICL. We hope that our study will bring researchers' attention to the limitations of MLLMs when dealing with perturbed inputs, and spur further investigations to address this issue. Our code and datasets are publicly available at https://github.com/NJUNLP/Hallu-PI.

cs.CV↗

arXiv2024

CFPFormer: Feature-pyramid like Transformer Decoder for Segmentation and Detection

Feature pyramids have been widely adopted in convolutional neural networks and transformers for tasks in medical image segmentation. However, existing models generally focus on the Encoder-side Transformer for feature extraction. We further explore the potential in improving the feature decoder with a well-designed architecture. We propose Cross Feature Pyramid Transformer decoder (CFPFormer), a novel decoder block that integrates feature pyramids and transformers. Even though transformer-like architecture impress with outstanding performance in segmentation, the concerns to reduce the redundancy and training costs still exist. Specifically, by leveraging patch embedding, cross-layer feature concatenation mechanisms, CFPFormer enhances feature extraction capabilities while complexity issue is mitigated by our Gaussian Attention. Benefiting from Transformer structure and U-shaped connections, our work is capable of capturing long-range dependencies and effectively up-sample feature maps. Experimental results are provided to evaluate CFPFormer on medical image segmentation datasets, demonstrating the efficacy and effectiveness. With a ResNet50 backbone, our method achieves 92.02\% Dice Score, highlighting the efficacy of our methods. Notably, our VGG-based model outperformed baselines with more complex ViT and Swin Transformer backbone.

cs.CV↗

arXiv2023

Taming the Exponential Action Set: Sublinear Regret and Fast Convergence to Nash Equilibrium in Online Congestion Games

The congestion game is a powerful model that encompasses a range of engineering systems such as traffic networks and resource allocation. It describes the behavior of a group of agents who share a common set of $F$ facilities and take actions as subsets with $k$ facilities. In this work, we study the online formulation of congestion games, where agents participate in the game repeatedly and observe feedback with randomness. We propose CongestEXP, a decentralized algorithm that applies the classic exponential weights method. By maintaining weights on the facility level, the regret bound of CongestEXP avoids the exponential dependence on the size of possible facility sets, i.e., $\binom{F}{k} \approx F^k$, and scales only linearly with $F$. Specifically, we show that CongestEXP attains a regret upper bound of $O(kF\sqrt{T})$ for every individual player, where $T$ is the time horizon. On the other hand, exploiting the exponential growth of weights enables CongestEXP to achieve a fast convergence rate. If a strict Nash equilibrium exists, we show that CongestEXP can converge to the strict Nash policy almost exponentially fast in $O(F\exp(-t^{1-\alpha}))$, where $t$ is the number of iterations and $\alpha \in (1/2, 1)$.

cs.GT↗