arXiv subjects

Kai Deng

Publications and source records attributed to Kai Deng.

At least 19 recordsLinked to original sources

arXiv2026

Auditing Semantic Gains in Sequential Recommendation: A Lightweight Recovery Test

Recent semantic and generative-retrieval recommenders report substantial improvements over ID-only sequential baselines, but it remains unclear whether these gains arise from language-model reasoning, semantic-ID generation, end-to-end semantic architectures, stronger offline item representations, or complementary semantic and collaborative signals. We investigate this attribution ambiguity through LIME-Rec, a lightweight and auditable recovery test. LIME-Rec combines three independent experts: a SASRec sequential expert, an ItemCF co-occurrence expert, and a semantic expert based on frozen BAAI/bge-base-en-v1.5 item embeddings. Their full-catalog scores are normalized per user and combined through auditable score-level fusion followed by bounded history calibration. The fusion gate and calibration head are fitted on validation data only, require no serving-time language-model inference, and keep each expert contribution separately inspectable. On Amazon Beauty, Toys, and Sports, LIME-Rec achieves R@10 scores of 0.0996, 0.1105, and 0.0593, outperforming the strongest comparison baseline by 7.0%-12.0%. Three-expert fusion without history calibration consistently outperforms calibrated SASRec, showing that calibration alone does not explain the recovery. Randomly permuting item-text embeddings across item IDs reduces R@10 by 13.6%-17.5%, indicating that the gains depend on genuine item-text correspondence rather than additional representation capacity. These results suggest that lightweight recovery from offline item representations and transparent fusion should be ruled out before improvements are attributed to serving-time language modeling, semantic-ID generation, or heavier semantic machinery.

cs.IR↗

arXiv2026

A Structure-Exploiting Implicit-Explicit Trust Region Method for Computing Second-Order Stationary Points of the Landau-Brazovskii Model

This work focuses on the reliable computation of second-order stationary points in the high-dimensional nonconvex energy landscape of the Landau-Brazovskii (LB) model, a fundamental model for studying phases and phase transitions. For this purpose, we develop an efficient implicit-explicit trust region (IMEX-TR) method. Trust region (TR) methods can avoid saddle-point stagnation and guarantee convergence to second-order stationary points under appropriate conditions. However, their direct application to the LB model has been impractical because the Hessian is dense if treated directly. The proposed IMEX-TR method overcomes this difficulty by exploiting the Hessian's special structure: the linear interaction part is diagonal in reciprocal space, whereas the nonlinear bulk-energy part is diagonal in physical space. Based on this structure, we design an efficient solver for the TR subproblem that with globally convergent guarantee and enjoys FFT-based acceleration, with $\mathcal O(N \log N)$ complexity per iteration. Existing first-order gradient-based methods for the LB model only guarantee convergence to first-order stationary points and may stagnate at saddle points. In contrast, the proposed IMEX-TR method inherits the theoretical guarantee of converging to second-order stationary points while remaining computationally practical. Numerical experiments verify the theoretical properties of the algorithm and demonstrate its robustness in locating stable phases from different initial conditions. Numerical results also show that IMEX-TR can escape unstable stationary states reached by first-order schemes and converge to physically meaningful second-order stationary points. These results suggest that targeting second-order stationary points provides an effective computational paradigm for exploring complex free-energy landscapes and identifying stable or metastable states.

math.NA↗

arXiv2026

Mitigating Hallucination in Financial Retrieval-Augmented Generation via Fine-Grained Knowledge Verification

In financial Retrieval-Augmented Generation (RAG) systems, models frequently rely on retrieved documents to generate accurate responses due to the time-sensitive nature of the financial domain. While retrieved documents help address knowledge gaps, model-generated responses still suffer from hallucinations that contradict the retrieved information. To mitigate this inconsistency, we propose a Reinforcement Learning framework enhanced with Fine-grained Knowledge Verification (RLFKV). Our method decomposes financial responses into atomic knowledge units and assesses the correctness of each unit to compute the fine-grained faithful reward. This reward offers more precise optimization signals, thereby improving alignment with the retrieved documents. Additionally, to prevent reward hacking (e.g., overly concise replies), we incorporate an informativeness reward that encourages the policy model to retain at least as many knowledge units as the base model. Experiments conducted on the public Financial Data Description (FDD) task and our newly proposed FDD-ANT dataset demonstrate consistent improvements, confirming the effectiveness of our approach.

cs.AI↗

arXiv2025

Antichiral surface states and higher-order topological states based on a modified Haldane model

Antichiral surface states, characterized by unidirectional propagation on parallel surfaces, offer unique potential for controlling classical waves. However, their realization typically relies on complex implementations of the two-dimensional modified Haldane model, limiting practical applications. Here, we propose a simplified scheme to realize such states within the nodal-line semimetal phase of a single-layer honeycomb lattice, by emulating the essential physics of the modified Haldane model through an introduced layer degree of freedom. Furthermore, we demonstrate that unequal vertical interlayer couplings can generate valley higher-order topological partial bandgaps, hosting coexisting one-dimensional hinge states and gapped antichiral surface states. We numerically verify these multiple topological states in acoustic crystals, establishing a versatile platform for advanced wave manipulation.

physics.app-ph↗

arXiv2025

FLAIR: Feedback Learning for Adaptive Information Retrieval

Recent advances in Large Language Models (LLMs) have driven the adoption of copilots in complex technical scenarios, underscoring the growing need for specialized information retrieval solutions. In this paper, we introduce FLAIR, a lightweight, feedback learning framework that adapts copilot systems' retrieval strategies by integrating domain-specific expert feedback. FLAIR operates in two stages: an offline phase obtains indicators from (1) user feedback and (2) questions synthesized from documentation, storing these indicators in a decentralized manner. An online phase then employs a two-track ranking mechanism to combine raw similarity scores with the collected indicators. This iterative setup refines retrieval performance for any query. Extensive real-world evaluations of FLAIR demonstrate significant performance gains on both previously seen and unseen queries, surpassing state-of-the-art approaches. The system has been successfully integrated into Copilot DECO, serving thousands of users at Microsoft, demonstrating its scalability and effectiveness in operational environments.

cs.IR↗

arXiv2025

VGGT-Long: Chunk it, Loop it, Align it -- Pushing VGGT's Limits on Kilometer-scale Long RGB Sequences

Foundation models for 3D vision have recently demonstrated remarkable capabilities in 3D perception. However, extending these models to large-scale RGB stream 3D reconstruction remains challenging due to memory limitations. In this work, we propose VGGT-Long, a simple yet effective system that pushes the limits of monocular 3D reconstruction to kilometer-scale, unbounded outdoor environments. Our approach addresses the scalability bottlenecks of existing models through a chunk-based processing strategy combined with overlapping alignment and lightweight loop closure optimization. Without requiring camera calibration, depth supervision or model retraining, VGGT-Long achieves trajectory and reconstruction performance comparable to traditional methods. We evaluate our method on KITTI, Waymo, and Virtual KITTI datasets. VGGT-Long not only runs successfully on long RGB sequences where foundation models typically fail, but also produces accurate and consistent geometry across various conditions. Our results highlight the potential of leveraging foundation models for scalable monocular 3D scene in real-world settings, especially for autonomous driving scenarios. Code is available at https://github.com/DengKaiCQ/VGGT-Long.

cs.CV↗

arXiv2025

AD-GS: Object-Aware B-Spline Gaussian Splatting for Self-Supervised Autonomous Driving

Modeling and rendering dynamic urban driving scenes is crucial for self-driving simulation. Current high-quality methods typically rely on costly manual object tracklet annotations, while self-supervised approaches fail to capture dynamic object motions accurately and decompose scenes properly, resulting in rendering artifacts. We introduce AD-GS, a novel self-supervised framework for high-quality free-viewpoint rendering of driving scenes from a single log. At its core is a novel learnable motion model that integrates locality-aware B-spline curves with global-aware trigonometric functions, enabling flexible yet precise dynamic object modeling. Rather than requiring comprehensive semantic labeling, AD-GS automatically segments scenes into objects and background with the simplified pseudo 2D segmentation, representing objects using dynamic Gaussians and bidirectional temporal visibility masks. Further, our model incorporates visibility reasoning and physically rigid regularization to enhance robustness. Extensive evaluations demonstrate that our annotation-free model significantly outperforms current state-of-the-art annotation-free methods and is competitive with annotation-dependent approaches.

cs.CV↗

arXiv2025

Evaluating the Effectiveness of Large Language Models in Solving Simple Programming Tasks: A User-Centered Study

As large language models (LLMs) become more common in educational tools and programming environments, questions arise about how these systems should interact with users. This study investigates how different interaction styles with ChatGPT-4o (passive, proactive, and collaborative) affect user performance on simple programming tasks. I conducted a within-subjects experiment where fifteen high school students participated, completing three problems under three distinct versions of the model. Each version was designed to represent a specific style of AI support: responding only when asked, offering suggestions automatically, or engaging the user in back-and-forth dialogue.Quantitative analysis revealed that the collaborative interaction style significantly improved task completion time compared to the passive and proactive conditions. Participants also reported higher satisfaction and perceived helpfulness when working with the collaborative version. These findings suggest that the way an LLM communicates, how it guides, prompts, and responds, can meaningfully impact learning and performance. This research highlights the importance of designing LLMs that go beyond functional correctness to support more interactive, adaptive, and user-centered experiences, especially for novice programmers.

cs.HC↗

arXiv2025

On the maximal matchings of trees

An independent edge set of graph $G$ is a matching, and is maximal if it is not a proper subset of any other matching of $G$. The number of all the maximal matchings of $G$ is denoted by $\Psi(G)$. In this paper, an algorithm to count $\Psi(T)$ for a tree $T$ is given. We show that for any tree $T$ with $n$ vertices, $\Psi(T)\geq\lceil\frac{n}{2}\rceil$, and the tree which obtained the lower bound is characterized.

math.CO↗

arXiv2025

GigaSLAM: Large-Scale Monocular SLAM with Hierarchical Gaussian Splats

Tracking and mapping in large-scale, unbounded outdoor environments using only monocular RGB input presents substantial challenges for existing SLAM systems. Traditional Neural Radiance Fields (NeRF) and 3D Gaussian Splatting (3DGS) SLAM methods are typically limited to small, bounded indoor settings. To overcome these challenges, we introduce GigaSLAM, the first RGB NeRF / 3DGS-based SLAM framework for kilometer-scale outdoor environments, as demonstrated on the KITTI, KITTI 360, 4 Seasons and A2D2 datasets. Our approach employs a hierarchical sparse voxel map representation, where Gaussians are decoded by neural networks at multiple levels of detail. This design enables efficient, scalable mapping and high-fidelity viewpoint rendering across expansive, unbounded scenes. For front-end tracking, GigaSLAM utilizes a metric depth model combined with epipolar geometry and PnP algorithms to accurately estimate poses, while incorporating a Bag-of-Words-based loop closure mechanism to maintain robust alignment over long trajectories. Consequently, GigaSLAM delivers high-precision tracking and visually faithful rendering on urban outdoor benchmarks, establishing a robust SLAM solution for large-scale, long-term scenarios, and significantly extending the applicability of Gaussian Splatting SLAM systems to unbounded outdoor environments. GitHub: https://github.com/DengKaiCQ/GigaSLAM.

cs.RO↗

arXiv2024

ENCO: Life-Cycle Management of Enterprise-Grade Copilots

Software engineers frequently grapple with the challenge of accessing disparate documentation and telemetry data, including TroubleShooting Guides (TSGs), incident reports, code repositories, and various internal tools developed by multiple stakeholders. While on-call duties are inevitable, incident resolution becomes even more daunting due to the obscurity of legacy sources and the pressures of strict time constraints. To enhance the efficiency of on-call engineers (OCEs) and streamline their daily workflows, we introduced DECO-a comprehensive framework for developing, deploying, and managing enterprise-grade copilots tailored to improve productivity in engineering routines. This paper details the design and implementation of the DECO framework, emphasizing its innovative NL2SearchQuery functionality and a lightweight agentic framework. These features support efficient and customized retrieval-augmented-generation (RAG) algorithms that not only extract relevant information from diverse sources but also select the most pertinent skills in response to user queries. This enables the addressing of complex technical questions and provides seamless, automated access to internal resources. Additionally, DECO incorporates a robust mechanism for converting unstructured incident logs into user-friendly, structured guides, effectively bridging the documentation gap. Since its launch in September 2023, ENCO has demonstrated its effectiveness through widespread adoption, enabling tens of thousands of interactions and engaging hundreds of monthly active users (MAU) across dozens of organizations within the company.

cs.SE↗

arXiv2022

Envelopes and principal component regression

Envelope methods offer targeted dimension reduction for various models. The overarching goal is to improve efficiency in multivariate parameter estimation by projecting the data onto a lower-dimensional subspace known as the envelope. Envelope approaches have advantages in analyzing data with highly correlated variables, but their iterative Grassmannian optimization algorithms do not scale very well with ultra high-dimensional data. While the connections between envelopes and partial least squares in multivariate linear regression have promoted recent progress in high-dimensional studies of envelopes, we propose a more straightforward way of envelope modeling from a novel principal components regression perspective. The proposed procedure, Non-Iterative Envelope Component Estimation (NIECE), has excellent computational advantages over the iterative Grassmannian optimization alternatives in high dimensions. We develop a unified NIECE theory that bridges the gap between envelope methods and principal components in regression. The new theoretical insights also shed light on the envelope subspace estimation error as a function of eigenvalue gaps of two symmetric positive definite matrices used in envelope modeling. We apply the new theory and algorithm to several envelope models, including response and predictor reduction in multivariate linear models, logistic regression, and Cox proportional hazard model. Simulations and illustrative data analysis show the potential for NIECE to improve standard methods in linear and generalized linear models significantly.

stat.ME↗

arXiv2021

Symmetric properties and two variants of shuffle-cubes

Li et al. in [Inf. Process. Lett. 77 (2001) 35--41] proposed the shuffle cube $SQ_{n}$ as an attractive interconnection network topology for massive parallel and distributed systems. By far, symmetric properties of the shuffle cube remains unknown. In this paper, we show that $SQ_{n}$ is not vertex-transitive for all $n>2$, which is not an appealing property in interconnection networks. To overcome this limitation, two novel vertex-transitive variants of the shuffle-cube, namely simplified shuffle-cube $SSQ_{n}$ and balanced shuffle cube $BSQ_{n}$ are introduced. Then, routing algorithms of $SSQ_{n}$ and $BSQ_{n}$ for all $n>2$ are given respectively. Furthermore, we show that both $SSQ_{n}$ and $BSQ_{n}$ possess Hamiltonian cycle embedding for all $n>2$. Finally, as a by-product, we mend a flaw in the Property 3 in [IEEE Trans. Comput. 46 (1997) 484--490].

math.CO↗

arXiv2021

A Doubly-Enhanced EM Algorithm for Model-Based Tensor Clustering

Modern scientific studies often collect data sets in the forms of tensors, which call for innovative statistical analysis methods. In particular, there is a pressing need for tensor clustering methods to understand the heterogeneity in the data. We propose a tensor normal mixture model (TNMM) approach to enable probabilistic interpretation and computational tractability. Our statistical model leverages the tensor covariance structure to reduce the number of parameters for parsimonious modeling, and at the same time explicitly exploits the correlations for better variable selection and clustering. We propose a doubly-enhanced expectation-maximization (DEEM) algorithm to perform clustering under this model. Both the E-step and the M-step are carefully tailored for tensor data in order to account for statistical accuracy and computational cost in high dimensions. Theoretical studies confirm that DEEM achieves consistent clustering even when the dimension of each mode of the tensors grows at an exponential rate of the sample size. Numerical studies demonstrate favorable performance of DEEM in comparison to existing methods.

stat.ME↗

arXiv2020

Forcing and anti-forcing polynomials of a polyomino graph

The forcing number of a perfect matching $M$ in a graph $G$ is the smallest number of edges inside $M$ that can not be contained in other perfect matchings. The anti-forcing number of $M$ is the smallest number of edges outside $M$ whose removal results in a subgraph with a single perfect matching, that is $M$. Recently, in order to investigate the distributions of forcing numbers and anti-forcing numbers, the forcing polynomial and anti-forcing polynomial were proposed, respectively. In this work, the forcing and anti-forcing polynomials of a polyomino graph are obtained. As consequences, the forcing and anti-forcing spectra of this polyomino graph are determined, and the asymptotic behaviors on the degree of freedom and the sum of all anti-forcing numbers are revealed, respectively.

math.CO↗

arXiv2020

Learning Heatmap-Style Jigsaw Puzzles Provides Good Pretraining for 2D Human Pose Estimation

The target of 2D human pose estimation is to locate the keypoints of body parts from input 2D images. State-of-the-art methods for pose estimation usually construct pixel-wise heatmaps from keypoints as labels for learning convolution neural networks, which are usually initialized randomly or using classification models on ImageNet as their backbones. We note that 2D pose estimation task is highly dependent on the contextual relationship between image patches, thus we introduce a self-supervised method for pretraining 2D pose estimation networks. Specifically, we propose Heatmap-Style Jigsaw Puzzles (HSJP) problem as our pretext-task, whose target is to learn the location of each patch from an image composed of shuffled patches. During our pretraining process, we only use images of person instances in MS-COCO, rather than introducing extra and much larger ImageNet dataset. A heatmap-style label for patch location is designed and our learning process is in a non-contrastive way. The weights learned by HSJP pretext task are utilised as backbones of 2D human pose estimator, which are then finetuned on MS-COCO human keypoints dataset. With two popular and strong 2D human pose estimators, HRNet and SimpleBaseline, we evaluate mAP score on both MS-COCO validation and test-dev datasets. Our experiments show that downstream pose estimators with our self-supervised pretraining obtain much better performance than those trained from scratch, and are comparable to those using ImageNet classification models as their initial backbones.

cs.CV↗

arXiv2020

Forcing and anti-forcing polynomials of perfect matchings of a pyrene system

The forcing number of a perfect matching of a graph was introduced by Harary et al., which originated from Klein and Randić's ideal of innate degree of freedom of Kekulé structure in molecular graph. On the opposite side in some sense, Vukičević and Trinajstié proposed the anti-forcing number of a graph, afterwards Lei et al. generalized this idea to single perfect matching. Recently the forcing and anti-forcing polynomials of perfect matchings of a graph were proposed as counting polynomials for perfect matchings with the same forcing number and anti-forcing number respectively. In this paper, we obtain the explicit expressions of forcing and anti-forcing polynomials of a pyrene system. As consequences, the forcing and anti-forcing spectra of a pyrene system are determined.

math.CO↗

arXiv2016

Extremal anti-forcing numbers of perfect matchings of graphs

The anti-forcing number of a perfect matching $M$ of a graph $G$ is the minimal number of edges not in $M$ whose removal to make $M$ as a unique perfect matching of the resulting graph. The set of anti-forcing numbers of all perfect matchings of $G$ is the anti-forcing spectrum of $G$. In this paper, we characterize the plane elementary bipartite graph whose minimum anti-forcing number is one. We show that the maximum anti-forcing number of a graph is at most its cyclomatic number. In particular, we characterize the graphs with the maximum anti-forcing number achieving the upper bound, such extremal graphs are a class of plane bipartite graphs. Finally, we determine the anti-forcing spectrum of an even polygonal chain in linear time.

math.CO↗