arXiv subjects

Lin Wang

Publications and source records attributed to Lin Wang.

At least 91 records · Page 5Linked to original sources

arXiv2026

Revisiting Invariant Learning for Out-of-Domain Generalization on Multi-Site Mammogram Datasets

Achieving health equity in Artificial Intelligence (AI) requires diagnostic models that maintain reliability across diverse populations. However, breast cancer screening systems frequently suffer from domain overfitting, degrading significantly when deployed to varying demographics. While Invariant Learning algorithms aim to mitigate this by suppressing site-specific correlations, their efficacy in medical imaging remains underexplored. This study comprehensively evaluates domain generalization techniques for mammography. We constructed a multi-source training environment aggregating datasets from the United States (CBIS-DDSM, EMBED), Portugal (INbreast, BCDR), and Cyprus (BMCD). To assess global generalizability, we evaluated performance on unseen cohorts from Egypt (CDD-CESM) and Sweden (CSAW-CC). We benchmarked Invariant Risk Minimization (IRM) and Variance Risk Extrapolation (VREx) against a rigorously optimized Empirical Risk Minimization (ERM) baseline. Contrary to expectations, standard ERM consistently outperformed specialized invariant mechanisms on out-of-domain testing. While VREx showed potential in stabilizing attention maps, invariant objectives proved unstable and prone to underfitting. We conclude that engineering equitable AI is currently best served by maximizing multi-national data diversity rather than relying on complex algorithmic invariance.

cs.CV↗

arXiv2026

DiEC: Diffusion Embedded Clustering

Deep clustering methods typically rely on a single, well-defined representation for clustering. In contrast, pretrained diffusion models provide abundant and diverse multi-scale representations across network layers and noise timesteps. However, a key challenge is how to efficiently identify the most clustering-friendly representation in the layer*timestep space. To address this issue, we propose Diffusion Embedded Clustering (DiEC), an unsupervised framework that performs clustering by leveraging optimal intermediate representations from pretrained diffusion models. DiEC systematically evaluates the clusterability of representations along the trajectory of network depth and noise timesteps. Meanwhile, an unsupervised search strategy is designed for recognizing the Clustering-optimal Layer (COL) and Clustering-optimal Timestep (COT) in the layer*timestep space of pretrained diffusion models, aiming to promote clustering performance and reduce computational overhead. DiEC is fine-tuned primarily with a structure-preserving DEC-style KL-divergence objective at the fixed COL + COT, together with a random-timestep diffusion denoising objective to maintain the generative capability of the pretrained model. Without relying on augmentation-based consistency constraints or contrastive learning, DiEC achieves excellent clustering performance across multiple benchmark datasets.

cs.LG↗

arXiv2026

Cells on Autopilot: Adaptive Cell (Re)Selection via Reinforcement Learning

The widespread deployment of 5G networks, together with the coexistence of 4G/LTE networks, provides mobile devices a diverse set of candidate cells to connect to. However, associating mobile devices to cells to maximize overall network performance, a.k.a. cell (re)selection, remains a key challenge for mobile operators. Today, cell (re)selection parameters are typically configured manually based on operator experience and rarely adapted to dynamic network conditions. In this work, we ask: Can an agent automatically learn and adapt cell (re)selection parameters to consistently improve network performance? We present a reinforcement learning (RL)-based framework called CellPilot that adaptively tunes cell (re)selection parameters by learning spatiotemporal patterns of mobile network dynamics. Our study with real-world data demonstrates that even a lightweight RL agent can outperform conventional heuristic reconfigurations by up to 167%, while generalizing effectively across different network scenarios. These results indicate that data-driven approaches can significantly improve cell (re)selection configurations and enhance mobile network performance.

cs.NI↗

arXiv2026

A Curriculum Learning Approach to Reinforcement Learning: Leveraging RAG for Multimodal Question Answering

This paper describes the solutions of the Dianping-Trust-Safety team for the META CRAG-MM challenge. The challenge requires building a comprehensive retrieval-augmented generation system capable for multi-modal multi-turn question answering. The competition consists of three tasks: (1) answering questions using structured data retrieved from an image-based mock knowledge graph, (2) synthesizing information from both knowledge graphs and web search results, and (3) handling multi-turn conversations that require context understanding and information aggregation from multiple sources. For Task 1, our solution is based on the vision large language model, enhanced by supervised fine-tuning with knowledge distilled from GPT-4.1. We further applied curriculum learning strategies to guide reinforcement learning, resulting in improved answer accuracy and reduced hallucination. For Task 2 and Task 3, we additionally leveraged web search APIs to incorporate external knowledge, enabling the system to better handle complex queries and multi-turn conversations. Our approach achieved 1st place in Task 1 with a significant lead of 52.38%, and 3rd place in Task 3, demonstrating the effectiveness of the integration of curriculum learning with reinforcement learning in our training pipeline.

cs.AI↗

arXiv2025

Subgroup Identification and Individualized Treatment Policies: A Tutorial on the Hybrid Two-Stage Workflow

Patients in clinical studies often exhibit heterogeneous treatment effect (HTE). Classical subgroup analyses provide inferential tools to test for effect modification, while modern machine learning methods estimate the Conditional Average Treatment Effect (CATE) to enable individual level prediction. Each paradigm has limitations: inference focused approaches may sacrifice predictive utility, and prediction focused approaches often lack statistical guarantees. We present a hybrid two-stage workflow that integrates these perspectives. Stage 1 applies statistical inference to test whether credible treatment effect heterogeneity exists with the protection against spurious findings. Stage 2 translates heterogeneity evidence into individualized treatment policies, evaluated by cross fitted doubly robust (DR) metrics with Neyman-Pearson (NP) constraints on harm. We illustrate the workflow with working examples based on simulated data and a real ACTG 175 HIV trial. This tutorial provides practical implementation checklists and discusses links to sponsor oriented HTE workflows, offering a transparent and auditable pathway from heterogeneity assessment to individualized treatment policies.

stat.AP↗

arXiv2025

Video Event Reasoning and Prediction by Fusing World Knowledge from LLMs with Vision Foundation Models

Current video understanding models excel at recognizing "what" is happening but fall short in high-level cognitive tasks like causal reasoning and future prediction, a limitation rooted in their lack of commonsense world knowledge. To bridge this cognitive gap, we propose a novel framework that synergistically fuses a powerful Vision Foundation Model (VFM) for deep visual perception with a Large Language Model (LLM) serving as a knowledge-driven reasoning core. Our key technical innovation is a sophisticated fusion module, inspired by the Q-Former architecture, which distills complex spatiotemporal and object-centric visual features into a concise, language-aligned representation. This enables the LLM to effectively ground its inferential processes in direct visual evidence. The model is trained via a two-stage strategy, beginning with large-scale alignment pre-training on video-text data, followed by targeted instruction fine-tuning on a curated dataset designed to elicit advanced reasoning and prediction skills. Extensive experiments demonstrate that our model achieves state-of-the-art performance on multiple challenging benchmarks. Notably, it exhibits remarkable zero-shot generalization to unseen reasoning tasks, and our in-depth ablation studies validate the critical contribution of each architectural component. This work pushes the boundary of machine perception from simple recognition towards genuine cognitive understanding, paving the way for more intelligent and capable AI systems in robotics, human-computer interaction, and beyond.

cs.CV↗

arXiv2025

Leveraging Large Language Models for Rare Disease Named Entity Recognition

Named Entity Recognition (NER) in the rare disease domain poses unique challenges due to limited labeled data, semantic ambiguity between entity types, and long-tail distributions. In this study, we evaluate the capabilities of GPT-4o for rare disease NER under low-resource settings, using a range of prompt-based strategies including zero-shot prompting, few-shot in-context learning, retrieval-augmented generation (RAG), and task-level fine-tuning. We design a structured prompting framework that encodes domain-specific knowledge and disambiguation rules for four entity types. We further introduce two semantically guided few-shot example selection methods to improve in-context performance while reducing labeling effort. Experiments on the RareDis Corpus show that GPT-4o achieves competitive or superior performance compared to BioClinicalBERT, with task-level fine-tuning yielding the strongest performance among the evaluated approaches and improving upon the previously reported BioClinicalBERT baseline. Cost-performance analysis reveals that few-shot prompting delivers high returns at low token budgets. RAG provides limited overall gains but can improve recall for challenging entity types, especially signs and symptoms. An error taxonomy highlights common failure modes such as boundary drift and type confusion, suggesting opportunities for post-processing and hybrid refinement. Our results demonstrate that prompt-optimized LLMs can serve as effective, scalable alternatives to traditional supervised models in biomedical NER, particularly in rare disease applications where annotated data is scarce.

cs.CL↗

arXiv2025

Probabilistic Approaches to The Energy Equality in Forced Surface Quasi-Geostrophic Equations

We explore probabilistic approaches to the deterministic energy equality for the forced Surface Quasi-Geostrophic (SQG) equation on a torus. First, we prove the zero-noise dynamical large deviations for a corresponding stochastic SQG equation, where the lower bound matches the upper bound on a certain closure of the weak-strong uniqueness class for the deterministic forced SQG equation. Furthermore, we show that the energy equality for the deterministic SQG equation holds on arbitrary time-reversible subsets of the domain where we match the upper bound and the lower bound. Conversely, the violation of the deterministic energy equality breaks the lower bound of large deviations. These results extend the existing techniques in Gess, Heydecker, and the second author \cite{arXiv:2311.02223} to generalized Sobolev spaces with negative indices. Finally, we provide an analysis of the restricted quasi-potential and prove a conditional equivalence compared to the rate function of large deviations for the Gaussian distribution. This suggests a potential connection between non-Gaussian large deviations in equilibrium for the stochastic SQG equation and the open problem regarding the uniqueness of the deterministic SQG equation.

math.PR↗

arXiv2025

A Conditioned UNet for Music Source Separation

In this paper we propose a conditioned UNet for Music Source Separation (MSS). MSS is generally performed by multi-output neural networks, typically UNets, with each output representing a particular stem from a predefined instrument vocabulary. In contrast, conditioned MSS networks accept an audio query related to a stem of interest alongside the signal from which that stem is to be extracted. Thus, a strict vocabulary is not required and this enables more realistic tasks in MSS. The potential of conditioned approaches for such tasks has been somewhat hidden due to a lack of suitable data, an issue recently addressed with the MoisesDb dataset. A recent method, Banquet, employs this dataset with promising results seen on larger vocabularies. Banquet uses Bandsplit RNN rather than a UNet and the authors state that UNets should not be suitable for conditioned MSS. We counter this argument and propose QSCNet, a novel conditioned UNet for MSS that integrates network conditioning elements in the Sparse Compressed Network for MSS. We find QSCNet to outperform Banquet by over 1dB SNR on a couple of MSS tasks, while using less than half the number of parameters.

cs.SD↗

arXiv2025

Roadmap: 2D Materials for Quantum Technologies

Two-dimensional (2D) materials have emerged as a versatile and powerful platform for quantum technologies, offering atomic-scale control, strong quantum confinement, and seamless integration into heterogeneous device architectures. Their reduced dimensionality enables unique quantum phenomena, including optically addressable spin defects, tunable single-photon emitters, low-dimensional magnetism, gate-controlled superconductivity, and correlated states in Moiré superlattices. This Roadmap provides a comprehensive overview of recent progress and future directions in exploiting 2D materials for quantum sensing, computation, communication, and simulation. We survey advances spanning spin defects and quantum sensing, quantum emitters and nonlinear photonics, computational theory and data-driven discovery of quantum defects, spintronic and magnonic devices, cavity-engineered quantum materials, superconducting and hybrid quantum circuits, quantum dots, Moiré quantum simulators, and quantum communication platforms. Across these themes, we identify common challenges in defect control, coherence preservation, interfacial engineering, and scalable integration, alongside emerging opportunities driven by machine$-$learning$-$assisted design and integrated experiment$-$theory feedback loops. By connecting microscopic quantum states to mesoscopic excitations and macroscopic device architectures, this Roadmap outlines a materials-centric framework for integrating coherent quantum functionalities and positions 2D materials as foundational building blocks for next-generation quantum technologies.

quant-ph↗

arXiv2025

Herman's converse KAM mechanism revisited

In his celebrated counterexample to the KAM theorem, Herman introduced a perturbation of an integrable system consisting of two components: a hyperbolic term and a bump function. He also remarked that it was unclear whether the bump function was truly necessary. In this note, we prove that the bump function is indeed necessary when more natural hyperbolic perturbations are considered. The proof of this necessity relies on an improved Siegel--Brjuno estimate and a parameter-dependent renormalization of resonances within the direct KAM method.

math.DS↗

arXiv2025

Physics-Informed Spiking Neural Networks via Conservative Flux Quantization

Real-time, physically-consistent predictions on low-power edge devices is critical for the next generation embodied AI systems, yet it remains a major challenge. Physics-Informed Neural Networks (PINNs) combine data-driven learning with physics-based constraints to ensure the model's predictions are with underlying physical principles.However, PINNs are energy-intensive and struggle to strictly enforce physical conservation laws. Brain-inspired spiking neural networks (SNNs) have emerged as a promising solution for edge computing and real-time processing. However, naively converting PINNs to SNNs degrades physical fidelity and fails to address long-term generalization issues. To this end, this paper introduce a novel Physics-Informed Spiking Neural Network (PISNN) framework. Importantly, to ensure strict physical conservation, we design the Conservative Leaky Integrate-and-Fire (C-LIF) neuron, whose dynamics structurally guarantee local mass preservation. To achieve robust temporal generalization, we introduce a novel Conservative Flux Quantization (CFQ) strategy, which redefines neural spikes as discrete packets of physical flux. Our CFQ learns a time-invariant physical evolution operator, enabling the PISNN to become a general-purpose solver -- conservative-by-construction. Extensive experiments show that our PISNN excels on diverse benchmarks. For both the canonical 1D heat equation and the more challenging 2D Laplace's Equation, it accurately simulates the system dynamics while maintaining perfect mass conservation by design -- a feat that is challenging for conventional PINNs. This work establishes a robust framework for fusing the rigor of scientific computing with the efficiency of neuromorphic engineering, paving the way for complex, long-term, and energy-efficient physics predictions for intelligent systems.

cs.LG↗

arXiv2025

Bio-inspired Integrated Networking and Control for Large-Scale Swarm: A Hierarchical Co-design

Unmanned aerial vehicle (UAV) swarms encounter the challenge of high overhead due to both network management and formation control requirements. In this paper, we propose a Bio-inspired Integrated Networking and Control (BINC) scheme, enabling efficient formation management for swarms comprising thousands of UAVs. The scheme forms a two-layer hierarchical structure, where network clusters and formations share the same groups so that cross-cluster control is eliminated. For networking, we design a fused routing message together with control information to reduce overhead, and limit clusters' size to local two-hop topologies for fast command transmission. For controlling, we develop a hybrid bio-inspired control approach, including a pigeon-like leader-follower algorithm within formations under the consideration of cluster topology maintenance, and a starling-like algorithm among formations that helps to improve the ability of obstacle avoidance. We establish a simulation platform for UAV swarms with over 1000 nodes, and experimental results show that the proposed BINC scheme can achieve highly maneuverable swarm formation marching with significant reduction on communication overhead.

cs.NI↗

arXiv2025

Multi-Level Damage-Aware Graph Learning for Resilient UAV Swarm Networks

Unmanned aerial vehicle (UAV) swarm networks leverage resilient algorithms to restore connectivity from communication network split issues. However, existing graph learning-based approaches face over-aggregation and non-convergence problems caused by uneven and sparse topology under massive damage. In this paper, we propose a novel Multi-Level Damage-Aware (MLDA) Graph Learning algorithm to generate recovery solutions, explicitly utilizing information about destroyed nodes to guide the recovery process. The algorithm first employs a Multi-Branch Damage Attention (MBDA) module as a pre-processing step, focusing attention on the critical relationships between remaining nodes and destroyed nodes in the global topology. By expanding multi-hop neighbor receptive fields of nodes to those damaged areas, it effectively mitigating the initial sparsity and unevenness before graph learning commences. Second, a Dilated Graph Convolution Network (DGCN) is designed to perform convolution on the MBDA-processed bipartite graphs between remaining and destroyed nodes. The DGCN utilizes a specialized bipartite graph convolution operation to aggregate features and incorporates a residual-connected architecture to extend depth, directly generating the target locations for recovery. We theoretically proved the convergence of the proposed algorithm and the computational complexity is acceptable. Simulation results show that the proposed algorithm can guarantee the connectivity restoration with excellent scalability, while significantly expediting the recovery time and improving the topology uniformity after recovery.

cs.NI↗

arXiv2025

The qualitative behavior for biharmonic functions on open manifolds

For a complete noncompact Riemannian manifold with nonnegative Ricci curvature, we show that bounded biharmonic functions are constant and the space consists of biharmonic functions with polynomial growth of a fixed rate is finite dimensional. Also, we derive a Weyl type bound for this space. Finally, we present a finite dimensional result for a class of fourth-order operators on $\mathbb{R}^n$ satisfying certain coefficient conditions.

math.DG↗

arXiv2025

AutoPrompt: Automated Red-Teaming of Text-to-Image Models via LLM-Driven Adversarial Prompts

Despite rapid advancements in text-to-image (T2I) models, their safety mechanisms are vulnerable to adversarial prompts, which maliciously generate unsafe images. Current red-teaming methods for proactively assessing such vulnerabilities usually require white-box access to T2I models, and rely on inefficient per-prompt optimization, as well as inevitably generate semantically meaningless prompts easily blocked by filters. In this paper, we propose APT (AutoPrompT), a black-box framework that leverages large language models (LLMs) to automatically generate human-readable adversarial suffixes for benign prompts. We first introduce an alternating optimization-finetuning pipeline between adversarial suffix optimization and fine-tuning the LLM utilizing the optimized suffix. Furthermore, we integrates a dual-evasion strategy in optimization phase, enabling the bypass of both perplexity-based filter and blacklist word filter: (1) we constrain the LLM generating human-readable prompts through an auxiliary LLM perplexity scoring, which starkly contrasts with prior token-level gibberish, and (2) we also introduce banned-token penalties to suppress the explicit generation of banned-tokens in blacklist. Extensive experiments demonstrate the excellent red-teaming performance of our human-readable, filter-resistant adversarial prompts, as well as superior zero-shot transferability which enables instant adaptation to unseen prompts and exposes critical vulnerabilities even in commercial APIs (e.g., Leonardo.Ai.).

cs.CV↗

arXiv2025

Learning Event-guided Exposure-agnostic Video Frame Interpolation via Adaptive Feature Blending

Exposure-agnostic video frame interpolation (VFI) is a challenging task that aims to recover sharp, high-frame-rate videos from blurry, low-frame-rate inputs captured under unknown and dynamic exposure conditions. Event cameras are sensors with high temporal resolution, making them especially advantageous for this task. However, existing event-guided methods struggle to produce satisfactory results on severely low-frame-rate blurry videos due to the lack of temporal constraints. In this paper, we introduce a novel event-guided framework for exposure-agnostic VFI, addressing this limitation through two key components: a Target-adaptive Event Sampling (TES) and a Target-adaptive Importance Mapping (TIM). Specifically, TES samples events around the target timestamp and the unknown exposure time to better align them with the corresponding blurry frames. TIM then generates an importance map that considers the temporal proximity and spatial relevance of consecutive features to the target. Guided by this map, our framework adaptively blends consecutive features, allowing temporally aligned features to serve as the primary cues while spatially relevant ones offer complementary support. Extensive experiments on both synthetic and real-world datasets demonstrate the effectiveness of our approach in exposure-agnostic VFI scenarios.

eess.IV↗

arXiv2025

On Non-KAM Invariant Circles for Area-Preserving Twist Maps

In this note, we investigate the dynamics of invariant circles in area-preserving twist maps. The invariant circles under consideration lie beyond the applicability of classical KAM theory, as the perturbations involved exceed the scope of standard KAM methods. By integrating both constructive and non-constructive techniques, we establish several results on the existence and breakdown of such non-KAM invariant circles. These findings provide new evidence in support of affirmative answers to two open questions posed by Mather in 1998.

math.DS↗