arXiv subjects

Wenbin Liu

Publications and source records attributed to Wenbin Liu.

At least 19 recordsLinked to original sources

arXiv2026

MathCoPilot: An Interactive System for Human-AI Symbiotic Paradigm of Mathematical Research

Existing LLM-based theorem provers have achieved impressive results on formal mathematics benchmarks, yet they remain confined to acting as autonomous agents that prove a stated proposition. In this paper, we propose MathCoPilot, a human-in-the-loop system that embodies a new human--AI symbiotic paradigm for mathematical research, in which the mathematician steers the high-level mathematical direction while AI agents carry out the detailed formalization and proof work under continuous human guidance. MathCoPilot unifies three core capabilities: (1) an interactive workbench where the mathematician and AI agents collaborate through a living proof blueprint that decomposes a proof into navigable steps the human can directly inspect, direct, and refine; (2) automated proving skill orchestration with adaptive knowledge base search and Lean-integrated iterative verification; and (3) topic-driven paper retrieval and automated formalization into a verified Lean knowledge base. Using MathCoPilot, we systematically compare four state-of-the-art LLMs, including Gemini~3.1~Pro, GPT-5.4, and Claude~Opus~4.7, on a FormalMATH subset and on two real PDE theorems requiring deep domain expertise, evaluating their ability to produce verified Lean~4 proofs and to identify errors in deliberately incorrect proofs. Our results show that while current models can handle undergraduate-level problems with high success rates under favorable autoformalization conditions, substantial challenges remain for domain-specific theorems requiring genuine mathematical understanding.

cs.AI↗

arXiv2026

Token-Level Policy Optimization: Linking Group-Level Rewards to Token-Level Aggregation via Sequence-Level Likelihood

Group Relative Policy Optimization (GRPO) has significantly advanced the reasoning ability of large language models (LLMs), particularly in their mathemat ical reasoning performance. However, GRPO and related entropy regularization methods still struggle with token-level sparse-rewards, which is an inherent chal lenge in chain-of-thought (CoT) reasoning. These approaches often rely on undifferen tiated token-level entropy regularization, which easily leads to entropy collapse or model degradation under sparse token rewards. In this work, we propose TEPO, a novel token-level framework that (1) leverages sequence-level likelihood to link group-level rewards with individual tokens via token-level aggregation, and (2) introduces a token-level KL-Divergence mask constraint that targets tokens with positive advantages and decreasing entropy to mitigate abrupt policy updates. Experiments demonstrate that TEPO not only achieves state-of-the-art performance on mathematical reasoning benchmarks but also markedly enhances training stability, reducing convergence time by 50% compared with GRPO/DAPO.

cs.CL↗

arXiv2025

GeoBridge: A Semantic-Anchored Multi-View Foundation Model Bridging Images and Text for Geo-Localization

Cross-view geo-localization infers a location by retrieving geo-tagged reference images that visually correspond to a query image. However, the traditional satellite-centric paradigm limits robustness when high-resolution or up-to-date satellite imagery is unavailable. It further underexploits complementary cues across views (\eg, drone, satellite, and street) and modalities (\eg, language and image). To address these challenges, we propose GeoBridge, a novel model that performs bidirectional matching across views and supports language-to-image retrieval. Going beyond traditional satellite-centric formulations, GeoBridge builds on a novel semantic-anchor mechanism that bridges multi-view features through textual descriptions for robust, flexible localization. In support of this task, we construct GeoLoc, the first large-scale, cross-modal, and multi-view aligned dataset comprising over 50,000 pairs of drone, street-view panorama, and satellite images as well as their textual descriptions, collected from 36 countries, ensuring both geographic and semantic alignment. We performed broad evaluations across multiple tasks. Experiments confirm that GeoLoc pre-training markedly improves geo-location accuracy for GeoBridge while promoting cross-domain generalization and cross-modal knowledge transfer. Code, dataset, and pretrained models will be released at https://github.com/MiliLab/GeoBridge.

cs.CV↗

arXiv2025

Trustworthy and Fair SkinGPT-R1 for Democratizing Dermatological Reasoning across Diverse Ethnicities

The clinical translation of dermatological AI is hindered by opaque reasoning and systematic performance disparities across skin tones. Here we present SkinGPT-R1, a multimodal large language model that integrates chain-of-thought diagnostic reasoning with a fairness-aware mixture-of-experts architecture for interpretable and equitable skin disease diagnosis. Through parameter-efficient adaptation of a frozen reasoning backbone, SkinGPT-R1 generates structured diagnostic reports comprising visual findings, differential reasoning, and final diagnosis. Across seven external datasets spanning diverse pathologies and imaging conditions, SkinGPT-R1 achieves state-of-the-art accuracy on six benchmarks, including 82.50\% on a challenging 40-class long-tail classification task (+19.30\% over leading baselines). Blinded evaluation by five board-certified dermatologists on 1,000 phenotypically balanced cases yields a mean score of 3.6 out of 5, with the highest ratings in safety (3.8) and reasoning coherence (3.6), indicating that the generated rationales are clinically safe, logically grounded, and suitable for supporting diagnostic decision-making. Critically, SkinGPT-R1 mitigates algorithmic bias across the full Fitzpatrick spectrum, achieving a robust worst-group performance of 41.40\% on the Fitz17k benchmark and a five-fold relative improvement in lower-bound accuracy on the DDI dataset compared to standard multimodal baselines. These results establish a framework for trustworthy, fair, and explainable AI-assisted dermatological diagnosis.

cs.CV↗

arXiv2025

Token-Level Policy Optimization: Linking Group-Level Rewards to Token-Level Aggregation via Markov Likelihood

Group Relative Policy Optimization (GRPO) has significantly advanced the reasoning ability of large language models (LLMs), particularly by boosting their mathematical performance. However, GRPO and related entropy-regularization methods still face challenges rooted in the sparse token rewards inherent to chain-of-thought (CoT). Current approaches often rely on undifferentiated token-level entropy adjustments, which frequently lead to entropy collapse or model collapse. In this work, we propose TEPO, a novel token-level framework that incorporates Markov Likelihood (sequence likelihood) links group-level rewards with tokens via token-level aggregation. Experiments show that TEPO consistently outperforms existing baselines across key metrics (including @k and accuracy). It not only sets a new state of the art on mathematical reasoning tasks but also significantly enhances training stability.

cs.CL↗

arXiv2025

Increasing the density limit with ECRH-assisted Ohmic start-up on EAST

High plasma density operation is crucial for a tokamak to achieve energy breakeven and a burning plasma. However, there is often an empirical upper limit of electron density in tokamak operation, namely the Greenwald density limit $n_G$, above which tokamaks generally disrupt. Achieving high-density operations above the density limit has been a long-standing challenge in magnetic confinement fusion research. Here, we report experimental results on EAST tokamak achieving the line-averaged electron density in the range of 1.3 $n_G$ to 1.65 $n_G$,while the usual range in EAST is (0.8-1.0)$n_G$. This is performed with ECRH-assisted Ohmic start-up and a sufficiently high initial neutral density. This is motivated by and consistent with predictions of a recent plasma-wall self-organization (PWSO) theory, that increasing ECRH power or pre-filled gas pressure leads to lower plasma temperatures around divertor target and higher density limits. In addition, the experiments are shown to operate in the density-free regime predicted by the PWSO model. These results suggest a promising scheme for substantially increasing the density limit in tokamaks, a critical advancement toward achieving the burning plasma.

physics.plasm-ph↗

arXiv2024

Optimal error estimates of the stochastic parabolic optimal control problem with integral state constraint

In this paper, the optimal strong error estimates for stochastic parabolic optimal control problem with additive noise and integral state constraint are derived based on time-implicit and finite element discretization. The continuous and discrete first-order optimality conditions are deduced by constructing the Lagrange functional, which contains forward-backward stochastic parabolic equations and a variational equation. The fully discrete version of forward-backward stochastic parabolic equations is introduced as an auxiliary problem and the optimal strong convergence orders are estimated, which further allows the optimal a priori error estimates for control, state, adjoint state and multiplier to be derived. Then, a simple and yet efficient gradient projection algorithm is proposed to solve stochastic parabolic control problem and its convergence rate is proved. Numerical experiments are carried out to illustrate the theoretical findings.

math.OC↗

arXiv2024

An efficient gradient projection method for stochastic optimal control problem with expected integral state constraint

In this work, we present an efficient gradient projection method for solving a class of stochastic optimal control problem with expected integral state constraint. The first order optimality condition system consisting of forward-backward stochastic differential equations and a variational equation is first derived. Then, an efficient gradient projection method with linear drift coefficient is proposed where the state constraint is guaranteed by constructing specific multiplier. Further, the Euler method is used to discretize the forward-backward stochastic differential equations and the associated conditional expectations are approximated by the least square Monte Carlo method, yielding the fully discrete iterative scheme. Error estimates of control and multiplier are presented, showing that the method admits first order convergence. Finally we present numerical examples to support the theoretical findings.

math.OC↗

arXiv2024

From Pixels to Gigapixels: Bridging Local Inductive Bias and Long-Range Dependencies with Pixel-Mamba

Histopathology plays a critical role in medical diagnostics, with whole slide images (WSIs) offering valuable insights that directly influence clinical decision-making. However, the large size and complexity of WSIs may pose significant challenges for deep learning models, in both computational efficiency and effective representation learning. In this work, we introduce Pixel-Mamba, a novel deep learning architecture designed to efficiently handle gigapixel WSIs. Pixel-Mamba leverages the Mamba module, a state-space model (SSM) with linear memory complexity, and incorporates local inductive biases through progressively expanding tokens, akin to convolutional neural networks. This enables Pixel-Mamba to hierarchically combine both local and global information while efficiently addressing computational challenges. Remarkably, Pixel-Mamba achieves or even surpasses the quantitative performance of state-of-the-art (SOTA) foundation models that were pretrained on millions of WSIs or WSI-text pairs, in a range of tumor staging and survival analysis tasks, {\bf even without requiring any pathology-specific pretraining}. Extensive experiments demonstrate the efficacy of Pixel-Mamba as a powerful and efficient framework for end-to-end WSI analysis.

cs.CV↗

arXiv2024

From Histopathology Images to Cell Clouds: Learning Slide Representations with Hierarchical Cell Transformer

It is clinically crucial and potentially very beneficial to be able to analyze and model directly the spatial distributions of cells in histopathology whole slide images (WSI). However, most existing WSI datasets lack cell-level annotations, owing to the extremely high cost over giga-pixel images. Thus, it remains an open question whether deep learning models can directly and effectively analyze WSIs from the semantic aspect of cell distributions. In this work, we construct a large-scale WSI dataset with more than 5 billion cell-level annotations, termed WSI-Cell5B, and a novel hierarchical Cell Cloud Transformer (CCFormer) to tackle these challenges. WSI-Cell5B is based on 6,998 WSIs of 11 cancers from The Cancer Genome Atlas Program, and all WSIs are annotated per cell by coordinates and types. To the best of our knowledge, WSI-Cell5B is the first WSI-level large-scale dataset integrating cell-level annotations. On the other hand, CCFormer formulates the collection of cells in each WSI as a cell cloud and models cell spatial distribution. Specifically, Neighboring Information Embedding (NIE) is proposed to characterize the distribution of cells within the neighborhood of each cell, and a novel Hierarchical Spatial Perception (HSP) module is proposed to learn the spatial relationship among cells in a bottom-up manner. The clinical analysis indicates that WSI-Cell5B can be used to design clinical evaluation metrics based on counting cells that effectively assess the survival risk of patients. Extensive experiments on survival prediction and cancer staging show that learning from cell spatial distribution alone can already achieve state-of-the-art (SOTA) performance, i.e., CCFormer strongly outperforms other competing methods.

cs.CV↗

arXiv2024

From Incomplete Coarse-Grained to Complete Fine-Grained: A Two-Stage Framework for Spatiotemporal Data Reconstruction

With the rapid development of various sensing devices, spatiotemporal data is becoming increasingly important nowadays. However, due to sensing costs and privacy concerns, the collected data is often incomplete and coarse-grained, limiting its application to specific tasks. To address this, we propose a new task called spatiotemporal data reconstruction, which aims to infer complete and fine-grained data from sparse and coarse-grained observations. To achieve this, we introduce a two-stage data inference framework, DiffRecon, grounded in the Denoising Diffusion Probabilistic Model (DDPM). In the first stage, we present Diffusion-C, a diffusion model augmented by ST-PointFormer, a powerful encoder designed to leverage the spatial correlations between sparse data points. Following this, the second stage introduces Diffusion-F, which incorporates the proposed T-PatternNet to capture the temporal pattern within sequential data. Together, these two stages form an end-to-end framework capable of inferring complete, fine-grained data from incomplete and coarse-grained observations. We conducted experiments on multiple real-world datasets to demonstrate the superiority of our method.

cs.LG↗

arXiv2024

UdeerLID+: Integrating LiDAR, Image, and Relative Depth with Semi-Supervised

Road segmentation is a critical task for autonomous driving systems, requiring accurate and robust methods to classify road surfaces from various environmental data. Our work introduces an innovative approach that integrates LiDAR point cloud data, visual image, and relative depth maps derived from images. The integration of multiple data sources in road segmentation presents both opportunities and challenges. One of the primary challenges is the scarcity of large-scale, accurately labeled datasets that are necessary for training robust deep learning models. To address this, we have developed the [UdeerLID+] framework under a semi-supervised learning paradigm. Experiments results on KITTI datasets validate the superior performance.

cs.CV↗

arXiv2024

End-to-end Multi-source Visual Prompt Tuning for Survival Analysis in Whole Slide Images

Survival analysis using pathology images poses a considerable challenge, as it requires the localization of relevant information from the multitude of tiles within whole slide images (WSIs). Current methods typically resort to a two-stage approach, where a pre-trained network extracts features from tiles, which are then used by survival models. This process, however, does not optimize the survival models in an end-to-end manner, and the pre-extracted features may not be ideally suited for survival prediction. To address this limitation, we present a novel end-to-end Visual Prompt Tuning framework for survival analysis, named VPTSurv. VPTSurv refines feature embeddings through an efficient encoder-decoder framework. The encoder remains fixed while the framework introduces tunable visual prompts and adaptors, thus permitting end-to-end training specifically for survival prediction by optimizing only the lightweight adaptors and the decoder. Moreover, the versatile VPTSurv framework accommodates multi-source information as prompts, thereby enriching the survival model. VPTSurv achieves substantial increases of 8.7% and 12.5% in the C-index on two immunohistochemical pathology image datasets. These significant improvements highlight the transformative potential of the end-to-end VPT framework over traditional two-stage methods.

eess.IV↗

arXiv2024

Toward Time-Continuous Data Inference in Sparse Urban CrowdSensing

Mobile Crowd Sensing (MCS) is a promising paradigm that leverages mobile users and their smart portable devices to perform various real-world tasks. However, due to budget constraints and the inaccessibility of certain areas, Sparse MCS has emerged as a more practical alternative, collecting data from a limited number of target subareas and utilizing inference algorithms to complete the full sensing map. While existing approaches typically assume a time-discrete setting with data remaining constant within each sensing cycle, this simplification can introduce significant errors, especially when dealing with long cycles, as real-world sensing data often changes continuously. In this paper, we go from fine-grained completion, i.e., the subdivision of sensing cycles into minimal time units, towards a more accurate, time-continuous completion. We first introduce Deep Matrix Factorization (DMF) as a neural network-enabled framework and enhance it with a Recurrent Neural Network (RNN-DMF) to capture temporal correlations in these finer time slices. To further deal with the continuous data, we propose TIME-DMF, which captures temporal information across unequal intervals, enabling time-continuous completion. Additionally, we present the Query-Generate (Q-G) strategy within TIME-DMF to model the infinite states of continuous data. Extensive experiments across five types of sensing tasks demonstrate the effectiveness of our models and the advantages of time-continuous completion.

cs.LG↗

arXiv2024

A generalized three-dimensional hybrid contact method for smoothed particle hydrodynamics

When the effects of relative motion at the solid object interfaces are not negligible, the contact method is required in the smoothed particle hydrodynamics (SPH) method to prevent virtual shear and tensile stresses. However, there is still a lack of a three-dimensional (3D) contact method that can be well applied to various deformation situations, especially for extreme deformation. In this study, we propose a generalized 3D hybrid contact method for SPH. First, an improved high accuracy free-surface particle detection method is developed, including optimization of the detection process to reduce the detection time and consideration of the effect of material compressibility on the filtering parameters to extend the existing semi-geometric method from the incompressible (weakly-compressible) field to the compressible field. Then, a novel 3D local surface reconstruction method is developed based on the free-surface particles and region growing method, including the selection of the initial edge, the principle of triangle expansion, and the evaluation function, followed by the surface-surface contact detection and enforcement of normal penalty and tangential friction forces according to the penalty function method and Coulomb friction law. Finally, the particle-particle contact method is added to deal with cases where surface-surface contact fails, e.g., some particles are unable to reconstruct the local surface when the material undergoes extreme deformations. The proposed method is validated by several numerical tests, and the results show that the proposed method is capable of handling various contact problems with accuracy and stability, including small, large, and extreme deformation problems.

physics.flu-dyn↗

arXiv2024

SkinCaRe: A Multimodal Dermatology Dataset Annotated with Medical Caption and Chain-of-Thought Reasoning

With the widespread application of artificial intelligence (AI), particularly deep learning (DL) and vision large language models (VLLMs), in skin disease diagnosis, the need for interpretability becomes crucial. However, existing dermatology datasets are limited in their inclusion of concept-level meta-labels, and none offer rich medical descriptions in natural language. This deficiency impedes the advancement of LLM-based methods in dermatologic diagnosis. To address this gap and provide a meticulously annotated dermatology dataset with comprehensive natural language descriptions, we introduce \textbf{SkinCaRe}, a comprehensive multimodal resource that unifies \textit{SkinCAP} and \textit{SkinCoT}. \textbf{SkinCAP} comprises 4,000 images sourced from the Fitzpatrick 17k skin disease dataset and the Diverse Dermatology Images dataset, annotated by board-certified dermatologists to provide extensive medical descriptions and captions. In addition, we introduce \textbf{SkinCoT}, a curated dataset pairing 3,041 dermatologic images with clinician-verified, hierarchical chain-of-thought (CoT) diagnoses. Each diagnostic narrative is rigorously evaluated against six quality criteria and iteratively refined until it meets a predefined standard of clinical accuracy and explanatory depth. Together, SkinCAP (captioning) and SkinCoT (reasoning), collectively referred to as SkinCaRe, encompass 7,041 expertly curated dermatologic cases and provide a unified and trustworthy resource for training multimodal models that both describe and explain dermatologic images. SkinCaRe is publicly available at https://huggingface.co/datasets/yuhos16/SkinCaRe.

cs.CV↗

arXiv2018

Non-cooperative game approach for task offloading in edge clouds

Task offloading provides a promising way to enhance the capability of the mobile terminal (also called terminal user) that is distributed on network edge and communicates edge clouds with wireless. Generally, there are multiple edge cloud nodes with distinct processing capability in a geographic area, which can offer computing service for various terminal users. Furthermore, the terminal users are competitive and selfish, i.e., each user takes into account only maximizing her own profit, while conducting task offloading strategies. In this paper, we focus on the resource management optimization for edge clouds, and formulate the problem of resource competition among terminal users as a non-cooperative game, in which the terminal user who acts as the player always pursues the minimization of the expected response time for her tasks by optimizing allocation strategies. We present the utility function of the user with queuing theory, and then prove the existence of Nash equilibrium for the formulated game. Using the concept of Nash bargaining solution to calculate the optimal task offloading scheme for the user, we propose a distributed task offloading algorithm with low computation complexity. The results of simulated experiments demonstrate that our method can quickly reach the Nash equilibrium point, and deliver satisfying performance at the expected response time of the user's tasks.

cs.GT↗

arXiv2018

Cell Selection with Deep Reinforcement Learning in Sparse Mobile Crowdsensing

Sparse Mobile CrowdSensing (MCS) is a novel MCS paradigm where data inference is incorporated into the MCS process for reducing sensing costs while its quality is guaranteed. Since the sensed data from different cells (sub-areas) of the target sensing area will probably lead to diverse levels of inference data quality, cell selection (i.e., choose which cells of the target area to collect sensed data from participants) is a critical issue that will impact the total amount of data that requires to be collected (i.e., data collection costs) for ensuring a certain level of quality. To address this issue, this paper proposes a Deep Reinforcement learning based Cell selection mechanism for Sparse MCS, called DR-Cell. First, we properly model the key concepts in reinforcement learning including state, action, and reward, and then propose to use a deep recurrent Q-network for learning the Q-function that can help decide which cell is a better choice under a certain state during cell selection. Furthermore, we leverage the transfer learning techniques to reduce the amount of data required for training the Q-function if there are multiple correlated MCS tasks that need to be conducted in the same target area. Experiments on various real-life sensing datasets verify the effectiveness of DR-Cell over the state-of-the-art cell selection mechanisms in Sparse MCS by reducing up to 15% of sensed cells with the same data inference quality guarantee.

cs.AI↗