Searcharxiv⌕ Search

arXiv subjects

Hong Chen

Publications and source records attributed to Hong Chen.

At least 55 records · Page 3Linked to original sources
arXiv2026

Human-AI Co-reasoning for Clinical Diagnosis with Evidence-Integrated Language Agent

We present PULSE, a medical reasoning agent that combines a domain-tuned large language model with scientific literature retrieval to support diagnostic decision-making in complex real-world cases. To evaluate its capabilities, we curated a benchmark of 82 authentic endocrinology case reports encompassing a broad spectrum of disease types and incidence levels. In controlled experiments, we compared PULSE's performance against physicians with varying levels of expertise-from residents to senior specialists-and examined how AI assistance influenced human diagnostic reasoning. PULSE attained expert-competitive accuracy, outperforming residents and junior specialists while matching senior specialist performance at both Top@1 and Top@4 thresholds. Unlike physicians, whose accuracy declined with disease rarity, PULSE maintained stable performance across incidence tiers. The agent also exhibited adaptive reasoning, increasing output length with case difficulty in a manner analogous to the longer deliberation observed among expert clinicians. When used collaboratively, PULSE enabled physicians to correct initial errors and broaden diagnostic hypotheses, but also introduced risks of automation bias. The study explores both serial and concurrent collaboration workflows, revealing that PULSE offers robust support across common and rare presentations. These findings underscore both the promise and the limitations of language model-based agents in clinical diagnosis, and offer a framework for evaluating their role in real-world decision-making.

cs.CL↗
arXiv2026

OralGPT-Plus: Learning to Use Visual Tools via Reinforcement Learning for Panoramic X-ray Analysis

Panoramic dental radiographs require fine-grained spatial reasoning, bilateral symmetry understanding, and multi-step diagnostic verification, yet existing vision-language models operate under a static single-pass paradigm that limits their clinical reliability. In this paper, we introduce OralGPT-Plus, an agentic vision-language model designed to perform iterative and symmetry-aware diagnostic reasoning for panoramic dental radiograph analysis. To support this paradigm, we construct DentalProbe, a five-thousand-image dataset with expert-curated diagnostic trajectories that provide structured supervision for localized inspection and contralateral comparison. We further develop a Reinspection-driven reinforcement learning framework that encourages clinically meaningful re-examination and stabilizes long-horizon reasoning with rubric-based reward and conditioned diagnostic-driven reward. In parallel, we present MMOral-X, the first benchmark for holistic panoramic diagnosis, containing 300 open-ended questions and region-level annotations across multiple difficulty levels. OralGPT-Plus demonstrates consistent and reliable improvements over strong baselines on MMOral-X and established panoramic benchmarks, indicating the effectiveness of interactive and symmetry-informed reasoning. Our work highlights the value of agentic modeling for dental imaging and provides a foundation for future research in clinically aligned panoramic radiograph analysis.

cs.CV↗
arXiv2026

LQRS: Learned Query Re-optimization Framework for Spark SQL

The query optimizer is a fundamental component of database management systems. Recent studies have shown that learned query optimizers outperform traditional cost-based query optimizers. However, they fail to exploit valuable runtime observations generated during query execution to dynamically re-optimize the plan, thereby limiting further improvements in query performance. To address this issue, we propose learned query re-optimization, which allows optimization decisions to be deferred to execution time and guided by actual runtime observations. We realize this idea through LQRS, a learned query re-optimization framework that builds upon Spark SQL, exploiting runtime observations for dynamic plan refinement. Specifically, LQRS employs a curriculum reinforcement learning strategy and jointly supports pre-execution and in-execution optimization, allowing knowledge learned during execution to directly benefit pre-execution planning. Furthermore, we design a plug-and-play planner extension built upon the extensibility interfaces of Spark SQL, enabling online plan modification. Experiments on Spark SQL demonstrate that LQRS reduces end-to-end execution time by up to 90% compared to other learned query optimizers and query re-optimization methods.

cs.DB↗
arXiv2026

Non-Hermiticity Induced Universal Anomalies in Kondo Conductance

Strong correlation, when combined with dissipation in open systems, can lead to a variety of exotic quantum phenomena. Here, we study nontrivial interplays between non-Fermi liquid behaviors emerging from strong correlation and non-Hermiticity arising from open systems. We propose a practical physical setup that realizes a non-Hermitian multichannel Kondo model. We identify a weak-coupling local moment fixed point and a strong-coupling non-Fermi liquid fixed point under PT symmetry, both are enriched by the non-Hermitian effect. Remarkably, universal unconventional Kondo conductance behaviors are found for both cases, which are distinct from all previously studied Kondo systems. Particularly, we show that an anomalous upturn of conductance could take place with increasing the temperature, originating from the interplay between non-Fermi liquid and non-Hermiticity. Our results identify a novel class of transport phenomena unrecognized before, driven by intertwined effects of correlation and dissipation.

cond-mat.str-el↗
arXiv2026

Beyond False Discovery Rate: A Stepdown Group SLOPE Approach for Grouped Variable Selection

High-dimensional feature selection is routinely required to balance statistical power with strict control of multiple-error metrics such as the k-Family-Wise Error Rate (k-FWER) and the False Discovery Proportion (FDP), yet some existing frameworks are confined to the narrower goal of controlling the expected False Discovery Rate (FDR) and can not exploit the group-structure of the covariates, such as Sorted L-One Penalized Estimation (SLOPE). We introduce the Group Stepdown SLOPE, a unified optimization procedure which is capable of embedding the Lehmann-Romano stepdown rules into SLOPE to achieve finite-sample guarantees under k-FWER and FDP thresholds. Specifically, we derive closed-form regularization sequences under orthogonal designs that provably bound k-FWER and FDP at user-specified levels, and extend these results to grouped settings via gk-SLOPE and gF-SLOPE, which control the analogous group-level errors gk-FWER and gFDP. For non-orthogonal general designs, we provide a calibrated data-driven sequence inspired by Gaussian approximation and Monte-Carlo correction, preserving convexity and scalability. Extensive simulations are conducted across sparse, correlated, and group-structured regimes. Empirical results corroborate our theoretical findings that the proposed methods achieve nominal error control, while yielding markedly higher power than competing stepdown procedures, thereby confirming the practical value of the theoretical advances.

stat.ME↗
arXiv2026

MemEmo: Evaluating Emotion in Memory Systems of Agents

Memory systems address the challenge of context loss in Large Language Model during prolonged interactions. However, compared to human cognition, the efficacy of these systems in processing emotion-related information remains inconclusive. To address this gap, we propose an emotion-enhanced memory evaluation benchmark to assess the performance of mainstream and state-of-the-art memory systems in handling affective information. We developed the \textbf{H}uman-\textbf{L}ike \textbf{M}emory \textbf{E}motion (\textbf{HLME}) dataset, which evaluates memory systems across three dimensions: emotional information extraction, emotional memory updating, and emotional memory question answering. Experimental results indicate that none of the evaluated systems achieve robust performance across all three tasks. Our findings provide an objective perspective on the current deficiencies of memory systems in processing emotional memories and suggest a new trajectory for future research and system optimization.

cs.CL↗
arXiv2026

DP-aware AdaLN-Zero: Taming Conditioning-Induced Heavy-Tailed Gradients in Differentially Private Diffusion

Condition injection enables diffusion models to generate context-aware outputs, which is essential for many time-series tasks. However, heterogeneous conditional contexts (e.g., observed history, missingness patterns or outlier covariates) can induce heavy-tailed per-example gradients. Under Differentially Private Stochastic Gradient Descent (DP-SGD), these rare conditioning-driven heavy-tailed gradients disproportionately trigger global clipping, resulting in outlier-dominated updates, larger clipping bias, and degraded utility under a fixed privacy budget. In this paper, we propose DP-aware AdaLN-Zero, a drop-in sensitivity-aware conditioning mechanism for conditional diffusion transformers that limits conditioning-induced gain without modifying the DP-SGD mechanism. DP-aware AdaLN-Zero jointly constrains conditioning representation magnitude and AdaLN modulation parameters via bounded re-parameterization, suppressing extreme gradient tail events before gradient clipping and noise injection. Empirically, DP-SGD equipped with DP-aware AdaLN-Zero improves interpolation/imputation and forecasting under matched privacy settings. We observe consistent gains on a real-world power dataset and two public ETT benchmarks over vanilla DP-SGD. Moreover, gradient diagnostics attribute these improvements to conditioning-specific tail reshaping and reduced clipping distortion, while preserving expressiveness in non-private training. Overall, these results show that sensitivity-aware conditioning can substantially improve private conditional diffusion training without sacrificing standard performance.

cs.LG↗
arXiv2026

Majorization via positivity of Jack and Macdonald polynomial differences

Majorization inequalities have a long history, going back to Maclaurin and Newton. They were recently studied for several families of symmetric functions, including by Cuttler--Greene--Skandera (2011), Sra (2016), Khare--Tao (2021), McSwiggen--Novak (2022), and Chen--Sahi (2024+) among others. Here we extend the inequalities by these authors to Jack and Macdonald polynomials, and obtain conjectural characterizations of majorization and of weak majorization of the underlying partitions. We prove these characterizations for two variables. In fact, we upgrade -- and prove in the above cases -- the characterization of majorization, to containment of Jack and Macdonald differences lying in the Muirhead semiring.

math.CO↗
arXiv2026

A Characterization of the Macdonald Hypergeometric Series ${}_rΦ_s(x;q,t)$ and ${}_rΦ_s(x,y;q,t)$ via $q$-Difference Equations

In two widely circulated manuscripts from the 1980s, I. G. Macdonald introduced certain multivariate hypergeometric series ${}_pF_q(x;α)$ and ${}_pF_q(x,y;α)$ and their $q$-analogs ${}_rΦ_s(x;q,t)$ and ${}_rΦ_s(x,y;q,t)$. These series are given by explicit expansions in Jack and Macdonald polynomials, and they generalize the hypergeometric functions of one and two matrix arguments from statistics. In a recent joint paper with Siddhartha Sahi, we constructed differential operators that characterize the Jack series ${}_pF_q$ thereby answering a question of Macdonald. In this paper we construct analogous $q$-difference operators that characterize the Macdonald series ${}_rΦ_s$. More precisely, we construct three $q$-difference operators $\mathcal A^{(x,y)}$, $\mathcal B^{(x)}$, $C^{(x)}$. The equation $\mathcal A^{(x,y)}(f(x,y))=0$ characterizes ${}_rΦ_s(x,y;q,t)$, while the equations $\mathcal B^{(x)}(f(x))=0$ and $\mathcal C^{(x)}(f(x))=0$ each characterize ${}_rΦ_s(x;q,t)$. These characterizations are subject to certain symmetry, boundary and stability condition. In the special case of ${}_2Φ_1(x;q,t)$, our operator $\mathcal B^{(x)}$ was previously constructed by Kaneko in 1996.

math.CO↗
arXiv2026

Anomalous Non-Hermitian Topological Anderson Insulator

Strong disorder drives conventional Hermitian systems into Anderson insulating states, suppressing all topological phases. Here, we unveil symmetry-protected, anomalous topological phases in the strong disorder limit of a non-Hermitian system, characterized by a scale-invariant merging of zero-energy modes. Using the maximally symmetric Jx lattice as an ideal platform and introducing specifically engineered (ABBA-type) symmetry-preserving non-Hermitian disorder, we observe a sequence of disorder-induced phase transitions: from a trivial insulator into and through a non-Hermitian topological Anderson insulator (TAI) phase, culminating in a stable anomalous non-Hermitian TAI phase characterized by a quantized polarization P_x \approx 0.25. Within this anomalous phase protected by the mobility gap, the zero-energy modes exhibit a distinct (N/2)-mode coalescence that scales with system size. Our findings demonstrate that non-Hermitian disorder engineered to preserve symmetry can induce and protect novel topological order inaccessible to conventional Hermitian disorder, thereby advancing the fundamental understanding of topological phenomena mediated by the interplay of disorder and non-Hermiticity.

cond-mat.mes-hall↗
arXiv2026

Dolphin-v2: Universal Document Parsing via Scalable Anchor Prompting

Document parsing has garnered widespread attention as vision-language models (VLMs) advance OCR capabilities. However, the field remains fragmented across dozens of specialized models with varying strengths, forcing users to navigate complex model selection and limiting system scalability. Moreover, existing two-stage approaches depend on axis-aligned bounding boxes for layout detection, failing to handle distorted or photographed documents effectively. To this end, we present Dolphin-v2, a two-stage document image parsing model that substantially improves upon the original Dolphin. In the first stage, Dolphin-v2 jointly performs document type classification (digital-born versus photographed) alongside layout analysis. For digital-born documents, it conducts finer-grained element detection with reading order prediction. In the second stage, we employ a hybrid parsing strategy: photographed documents are parsed holistically as complete pages to handle geometric distortions, while digital-born documents undergo element-wise parallel parsing guided by the detected layout anchors, enabling efficient content extraction. Compared with the original Dolphin, Dolphin-v2 introduces several crucial enhancements: (1) robust parsing of photographed documents via holistic page-level understanding, (2) finer-grained element detection (21 categories) with semantic attribute extraction such as author information and document metadata, and (3) code block recognition with indentation preservation, which existing systems typically lack. Comprehensive evaluations are conducted on DocPTBench, OmniDocBench, and our self-constructed RealDoc-160 benchmark. The results demonstrate substantial improvements: +14.78 points overall on the challenging OmniDocBench and 91% error reduction on photographed documents, while maintaining efficient inference through parallel processing.

cs.CV↗
arXiv2026

SONIC: Segmented Optimized Nexus for Information Compression in Key-Value Caching

The linear growth of Key-Value (KV) cache remains a bottleneck for multi-turn LLM deployment. Existing KV cache compression methods often fail to account for the structural properties of multi-turn dialogues, relying on heuristic eviction that risks losing critical context. We propose \textbf{SONIC}, a learning-based framework that compresses historical segments into compact and semantically rich \textbf{Nexus} tokens. By integrating dynamic budget training, SONIC allows flexible adaptation to varying memory constraints without retraining. Experiments show that at compression ratios of 80\% and 50\%, SONIC consistently outperforms baselines such as H2O and StreamingLLM on four diverse multi-turn benchmarks. Specifically, on the widely used MTBench101 benchmark, SONIC achieves an average score improvement of 35.55\% over state-of-the-art baselines, validating its effectiveness in sustaining coherent multi-turn dialogues. Furthermore, SONIC enhances deployment efficiency, accelerating the overall inference process by 50.1\% compared to full-context generation.

cs.CL↗
arXiv2026

CURVE: Learning Causality-Inspired Invariant Representations for Robust Scene Understanding via Uncertainty-Guided Regularization

Scene graphs provide structured abstractions for scene understanding, yet they often overfit to spurious correlations, severely hindering out-of-distribution generalization. To address this limitation, we propose CURVE, a causality-inspired framework that integrates variational uncertainty modeling with uncertainty-guided structural regularization to suppress high-variance, environment-specific relations. Specifically, we apply prototype-conditioned debiasing to disentangle invariant interaction dynamics from environment-dependent variations, promoting a sparse and domain-stable topology. Empirically, we evaluate CURVE in zero-shot transfer and low-data sim-to-real adaptation, verifying its ability to learn domain-stable sparse topologies and provide reliable uncertainty estimates to support risk prediction under distribution shifts.

cs.CV↗
arXiv2026

Realization of staircase topological Anderson phase transitions

One-dimensional topological Anderson insulators provide a paradigm for disorder-induced topological phases in which the underlying system turns from a trivial to a topological phase. It is widely recognized that the latter vanishes at large disorder amplitude. Here, and contrary to the general belief, we provide evidence for a successive disorder-driven topological transitions in a single-wall nanotube, culminating in a topological Anderson phase that remains unexpectedly robust at strong disorder. This phenomenon is confirmed by analysis of the corresponding topological invariant, which increases stepwise as disorder increases, giving evidence for the emergence of edge states. We experimentally implement these topological Anderson staircase phase transitions in a one-dimensional topolectrical circuit, where the persistence of edge states is revealed by node-voltage measurements. The robustness of the edge states is corroborated by numerical calculations of their localization properties. Our work opens the road to topological disordertronics, where topological phases can be tuned by disorder.

cond-mat.mes-hall↗
arXiv2026

Deconstructing Pre-training: Knowledge Attribution Analysis in MoE and Dense Models

Mixture-of-Experts (MoE) architectures decouple model capacity from per-token computation, enabling scaling beyond the computational limits imposed by dense scaling laws. Yet how MoE architectures shape knowledge acquisition during pre-training, and how this process differs from dense architectures, remains unknown. To address this issue, we introduce Gated-LPI (Log-Probability Increase), a neuron-level attribution metric that decomposes log-probability increase across neurons. We present a time-resolved comparison of knowledge acquisition dynamics in MoE and dense architectures, tracking checkpoints over 1.2M training steps (~ 5.0T tokens) and 600K training steps (~ 2.5T tokens), respectively. Our experiments uncover three patterns: (1) Low-entropy backbone. The top approximately 1% of MoE neurons capture over 45% of positive updates, forming a high-utility core, which is absent in the dense baseline. (2) Early consolidation. The MoE model locks into a stable importance profile within < 100K steps, whereas the dense model remains volatile throughout training. (3) Functional robustness. Masking the ten most important MoE attention heads reduces relational HIT@10 by < 10%, compared with > 50% for the dense model, showing that sparsity fosters distributed -- rather than brittle -- knowledge storage. These patterns collectively demonstrate that sparsity fosters an intrinsically stable and distributed computational backbone from early in training, helping bridge the gap between sparse architectures and training-time interpretability.

cs.AI↗
arXiv2026

TableCache: Primary Foreign Key Guided KV Cache Precomputation for Low Latency Text-to-SQL

In Text-to-SQL tasks, existing LLM-based methods often include extensive database schemas in prompts, leading to long context lengths and increased prefilling latency. While user queries typically focus on recurrent table sets-offering an opportunity for KV cache sharing across queries-current inference engines, such as SGLang and vLLM, generate redundant prefix cache copies when processing user queries with varying table orders. To address this inefficiency, we propose precomputing table representations as KV caches offline and querying the required ones online. A key aspect of our approach is the computation of table caches while preserving primary foreign key relationships between tables. Additionally, we construct a Table Trie structure to facilitate efficient KV cache lookups during inference. To enhance cache performance, we introduce a cache management system with a query reranking strategy to improve cache hit rates and a computation loading pipeline for parallelizing model inference and cache loading. Experimental results show that our proposed TableCache achieves up to a 3.62x speedup in Time to First Token (TTFT) with negligible performance degradation.

cs.CL↗
arXiv2026

A Fast Semidefinite Convex Relaxation for Optimal Control Problems With Spatio-Temporal Constraints

Solving optimal control problems (OCPs) of autonomous agents operating under spatial and temporal constraints fast and accurately is essential in applications ranging from eco-driving of autonomous vehicles to quadrotor navigation. However, the nonlinear programs approximating the OCPs are inherently nonconvex due to the coupling between the dynamics and the event timing, and therefore, they are challenging to solve. Most approaches address this challenge by predefining waypoint times or just using nonconvex trajectory optimization, which simplifies the problem but often yields suboptimal solutions. To significantly improve the numerical properties, we propose a formulation with a time-scaling direct multiple shooting scheme that partitions the prediction horizon into segments aligned with characteristic time constraints. Moreover, we develop a fast semidefinite-programming-based convex relaxation that exploits the sparsity pattern of the lifted formulation. Comprehensive simulation studies demonstrate the solution optimality and computational efficiency. Furthermore, real-world experiments on a quadrotor waypoint flight task with constrained open time windows validate the practical applicability of the approach in complex environments.

cs.RO↗
arXiv2025

D$^{2}$-VPR: A Parameter-efficient Visual-foundation-model-based Visual Place Recognition Method via Knowledge Distillation and Deformable Aggregation

Visual Place Recognition (VPR) aims to determine the geographic location of a query image by retrieving its most visually similar counterpart from a geo-tagged reference database. Recently, the emergence of the powerful visual foundation model, DINOv2, trained in a self-supervised manner on massive datasets, has significantly improved VPR performance. This improvement stems from DINOv2's exceptional feature generalization capabilities but is often accompanied by increased model complexity and computational overhead that impede deployment on resource-constrained devices. To address this challenge, we propose $D^{2}$-VPR, a $D$istillation- and $D$eformable-based framework that retains the strong feature extraction capabilities of visual foundation models while significantly reducing model parameters and achieving a more favorable performance-efficiency trade-off. Specifically, first, we employ a two-stage training strategy that integrates knowledge distillation and fine-tuning. Additionally, we introduce a Distillation Recovery Module (DRM) to better align the feature spaces between the teacher and student models, thereby minimizing knowledge transfer losses to the greatest extent possible. Second, we design a Top-Down-attention-based Deformable Aggregator (TDDA) that leverages global semantic features to dynamically and adaptively adjust the Regions of Interest (ROI) used for aggregation, thereby improving adaptability to irregular structures. Extensive experiments demonstrate that our method achieves competitive performance compared to state-of-the-art approaches. Meanwhile, it reduces the parameter count by approximately 64.2% (compared to CricaVPR).Code is available at https://github.com/tony19980810/D2VPR.

cs.CV↗