arXiv subjects

Ming-Liang Zhang

Publications and source records attributed to Ming-Liang Zhang.

14 recordsLinked to original sources

arXiv2026

Mobile-R1: Towards Interactive Capability for VLM-Based Mobile Agent via Systematic Training

Vision-language model-based mobile agents have gained the ability to understand complex instructions and mobile screenshots, benefiting from reinforcement learning paradigms like Group Relative Policy Optimization (GRPO). However, existing approaches centers on offline training or local action-level rewards often trap agents in local optima, hindering effective exploration and error correction with the environment. Crucially, we find that directly applying task-level rewards often leads to convergence difficulties due to the sparse nature of GUI interactions. To address these challenges, we present \textbf{Mobile-R1}, a systematic training recipe that bridges atomic action execution and strategic task completion. We propose a hierarchical curriculum consisting of three stages: (1) format alignment for reasoning structure, (2) on-policy exploration with verifiable action feedback to ground basic execution, and (3) multi-turn task-level training with realistic environment to unlock exploration and self-correction. This hierarchical strategy effectively bootstraps the agent, significantly enhancing its capability for exploration and self-correction (the ``Eureka'' moments). Furthermore, addressing the critical scarcity of diverse GUI data in non-English ecosystems, we contribute a comprehensive Chinese mobile dataset covering 28 applications with 24,521 high-quality manual annotations, and establish a rigorous benchmark with 500 trajectories. We will open source all resources, including the dataset, benchmark, model weight, and codes: https://mobile-r1.github.io/Mobile-R1/.

cs.AI↗

arXiv2026

Geoparsing: Diagram Parsing for Plane and Solid Geometry with a Unified Formal Language

Multimodal Large Language Models (MLLMs) have achieved remarkable progress but continue to struggle with geometric reasoning, primarily due to the perception bottleneck regarding fine-grained visual elements. While formal languages have aided plane geometry understanding, solid geometry which requires spatial understanding remains largely unexplored. In this paper, we address this challenge by designing a unified formal language that integrates plane and solid geometry, comprehensively covering geometric structures and semantic relations. We construct GDP-29K, a large-scale dataset comprising 20k plane and 9k solid geometry samples collected from diverse real-world sources, each paired with its ground-truth formal description. To ensure syntactic correctness and geometric consistency, we propose a training paradigm that combines Supervised Fine-Tuning with Reinforcement Learning via Verifiable Rewards. Experiments show that our approach achieves state-of-the-art parsing performance. Furthermore, we demonstrate that our parsed formal descriptions serve as a critical cognitive scaffold, significantly boosting MLLMs' capabilities for downstream geometry reasoning tasks. Our data and code are available at Geoparsing.

cs.CV↗

arXiv2025

From System 1 to System 2: A Survey of Reasoning Large Language Models

Achieving human-level intelligence requires refining the transition from the fast, intuitive System 1 to the slower, more deliberate System 2 reasoning. While System 1 excels in quick, heuristic decisions, System 2 relies on logical reasoning for more accurate judgments and reduced biases. Foundational Large Language Models (LLMs) excel at fast decision-making but lack the depth for complex reasoning, as they have not yet fully embraced the step-by-step analysis characteristic of true System 2 thinking. Recently, reasoning LLMs like OpenAI's o1/o3 and DeepSeek's R1 have demonstrated expert-level performance in fields such as mathematics and coding, closely mimicking the deliberate reasoning of System 2 and showcasing human-like cognitive abilities. This survey begins with a brief overview of the progress in foundational LLMs and the early development of System 2 technologies, exploring how their combination has paved the way for reasoning LLMs. Next, we discuss how to construct reasoning LLMs, analyzing their features, the core methods enabling advanced reasoning, and the evolution of various reasoning LLMs. Additionally, we provide an overview of reasoning benchmarks, offering an in-depth comparison of the performance of representative reasoning LLMs. Finally, we explore promising directions for advancing reasoning LLMs and maintain a real-time \href{https://github.com/zzli2022/Awesome-Slow-Reason-System}{GitHub Repository} to track the latest developments. We hope this survey will serve as a valuable resource to inspire innovation and drive progress in this rapidly evolving field.

cs.AI↗

arXiv2024

Fuse, Reason and Verify: Geometry Problem Solving with Parsed Clauses from Diagram

Geometry problem solving (GPS) requires capacities of multi-modal understanding, multi-hop reasoning and theorem knowledge application. In this paper, we propose a neural-symbolic model for plane geometry problem solving (PGPS), named PGPSNet-v2, with three key steps: modal fusion, reasoning process and knowledge verification. In modal fusion, we leverage textual clauses to express fine-grained structural and semantic content of geometry diagram, and fuse diagram with textual problem efficiently through structural-semantic pre-training. For reasoning, we design an explicable solution program to describe the geometric reasoning process, and employ a self-limited decoder to generate solution program autoregressively. To reduce solution errors, a multi-level theorem verifier is proposed to eliminate solutions that do not match geometric principles, alleviating the hallucination of the neural model. We also construct a large-scale geometry problem dataset called PGPS9K, containing fine-grained annotations of textual clauses, solution program and involved knowledge tuples. Extensive experiments on datasets Geometry3K and PGPS9K show that our PGPSNet solver outperforms existing symbolic and neural solvers in GPS performance, while maintaining good explainability and reliability, and the solver components (fusion, reasoning, verification) are all justified effective.

cs.AI↗

arXiv2024

CMMaTH: A Chinese Multi-modal Math Skill Evaluation Benchmark for Foundation Models

Due to the rapid advancements in multimodal large language models, evaluating their multimodal mathematical capabilities continues to receive wide attention. Despite the datasets like MathVista proposed benchmarks for assessing mathematical capabilities in multimodal scenarios, there is still a lack of corresponding evaluation tools and datasets for fine-grained assessment in the context of K12 education in Chinese language. To systematically evaluate the capability of multimodal large models in solving Chinese multimodal mathematical problems, we propose a Chinese Multi-modal Math Skill Evaluation Benchmark, named CMMaTH, contraining 23k multimodal K12 math related questions, forming the largest Chinese multimodal mathematical problem benchmark to date. CMMaTH questions from elementary to high school levels, provide increased diversity in problem types, solution objectives, visual elements, detailed knowledge points, and standard solution annotations. We have constructed an open-source tool GradeGPT integrated with the CMMaTH dataset, facilitating stable, rapid, and cost-free model evaluation. Our data and code are available.

cs.CL↗

arXiv2024

LANS: A Layout-Aware Neural Solver for Plane Geometry Problem

Geometry problem solving (GPS) is a challenging mathematical reasoning task requiring multi-modal understanding, fusion, and reasoning. Existing neural solvers take GPS as a vision-language task but are short in the representation of geometry diagrams that carry rich and complex layout information. In this paper, we propose a layout-aware neural solver named LANS, integrated with two new modules: multimodal layout-aware pre-trained language module (MLA-PLM) and layout-aware fusion attention (LA-FA). MLA-PLM adopts structural-semantic pre-training (SSP) to implement global relationship modeling, and point-match pre-training (PMP) to achieve alignment between visual points and textual points. LA-FA employs a layout-aware attention mask to realize point-guided cross-modal fusion for further boosting layout awareness of LANS. Extensive experiments on datasets Geometry3K and PGPS9K validate the effectiveness of the layout-aware modules and superior problem-solving performance of our LANS solver, over existing symbolic and neural solvers. The code will be made public available soon.

cs.CV↗

arXiv2023

A Multi-Modal Neural Geometric Solver with Textual Clauses Parsed from Diagram

Geometry problem solving (GPS) is a high-level mathematical reasoning requiring the capacities of multi-modal fusion and geometric knowledge application. Recently, neural solvers have shown great potential in GPS but still be short in diagram presentation and modal fusion. In this work, we convert diagrams into basic textual clauses to describe diagram features effectively, and propose a new neural solver called PGPSNet to fuse multi-modal information efficiently. Combining structural and semantic pre-training, data augmentation and self-limited decoding, PGPSNet is endowed with rich knowledge of geometry theorems and geometric representation, and therefore promotes geometric understanding and reasoning. In addition, to facilitate the research of GPS, we build a new large-scale and fine-annotated GPS dataset named PGPS9K, labeled with both fine-grained diagram annotation and interpretable solution program. Experiments on PGPS9K and an existing dataset Geometry3K validate the superiority of our method over the state-of-the-art neural solvers. Our code, dataset and appendix material are available at \url{https://github.com/mingliangzhang2018/PGPS}.

cs.AI↗

arXiv2022

Plane Geometry Diagram Parsing

Geometry diagram parsing plays a key role in geometry problem solving, wherein the primitive extraction and relation parsing remain challenging due to the complex layout and between-primitive relationship. In this paper, we propose a powerful diagram parser based on deep learning and graph reasoning. Specifically, a modified instance segmentation method is proposed to extract geometric primitives, and the graph neural network (GNN) is leveraged to realize relation parsing and primitive classification incorporating geometric features and prior knowledge. All the modules are integrated into an end-to-end model called PGDPNet to perform all the sub-tasks simultaneously. In addition, we build a new large-scale geometry diagram dataset named PGDP5K with primitive level annotations. Experiments on PGDP5K and an existing dataset IMP-Geometry3K show that our model outperforms state-of-the-art methods in four sub-tasks remarkably. Our code, dataset and appendix material are available at https://github.com/mingliangzhang2018/PGDP.

cs.CV↗

arXiv2018

Piezoelectric scattering of optical polaron in halide perovskites

For the intrinsic carriers of MAPbBr$_{3}$, the temperature $T$ dependent mobility $μ(T)$ of behaves like $μ\propto T^{-1/2}$ in piezoelectric tetragonal phase, $μ\propto T^{-1.4}$ in non-piezoelectric cubic phase. But for the photo-generated carriers in other halide perovskites ABX$_{3}$, $μ\propto T^{-3/2}$ behavior is typical. Due to the strong interaction of carrier with longitudinal optical phonon, in ABX$_{3}$ the carriers mainly exist as optical polarons. The softness of ABX$_{3}$ renders it without inversion center in tetragonal phase, which allows piezoelectric effect at low carrier concentration. The variations of $μ(T)$ behavior results from (1) the wave vector dependence of the piezoelectric interaction of polarons with acoustic phonons is different from that of ordinary polaron-acoustic phonon interaction; (2) the residual interaction of polaron with 2 longitudinal optical phonons can be ignored at low temperature, but is important at higher temperature; and (3) the concentration of intrinsic carriers is determined by temperature, while the concentration of photo-generated carriers is determined by the incident flux of photons.

cond-mat.mtrl-sci↗

arXiv2011

Approximate Theory of Temperature Coefficient of Resistivity of Amorphous Semiconductors

In this paper, we develop an approximate theory of the temperature coefficient of resistivity (TCR) and conductivity based upon the recently proposed Microscopic Response Method. By introducing suitable approximations for the lattice dynamics, localized and extended electronic states, we produce new explicit forms for the conductivity and TCR, which depend on easily accessible material parameters. The theory is in reasonable agreement with experiments on a-Si:H and a-Ge:H. A long-standing puzzle, a \textquotedblleft kink\textquotedblright\ in the experimental $% \log_{10}σ$ vs. 1/T curve, is predicted by the theory and attributed to localized to extended transitions, which have not been properly handled in earlier theories.

cond-mat.stat-mech↗

arXiv2011

Temperature Coefficient of Resistivity in Amorphous Semiconductors

By invoking the microscopic response method in conjunction with a reasonable set of approximations, we obtain new explicit expressions for the electrical conductivity and temperature coefficient of resistivity (TCR) in amorphous semiconductors, especially a-Si:H and a-Ge:H. The predicted TCR for n-doped a-Si:H and a-Ge:H is in agreement with experiments. The conductivity from the transitions from a localized state to an extended state (LE) is comparable to that from the transitions between two localized states (LL). This resolves a long-standing anomaly, a "kink" in the experimental $\log_{10}σ$ vs. T$^{-1}$ curve.

cond-mat.stat-mech↗

arXiv2010

Phonon driven transport in amorphous semiconductors: Transition probabilities

Starting from Holstein's work on small polaron hopping, the evolution equations for localized and extended states in the presence of atomic vibrations are systematically derived for an amorphous semiconductor. The transition probabilities are obtained for transitions between all combinations of localized and extended states. For any transition process involving a localized state, the activation energy is not simply the energy difference between the final and initial states; the reorganization energy of atomic configuration is also included as an important part of the activation energy (Marcus form). The activation energy for the transitions between localized states decreases with rising temperature and leads to the Meyer-Neldel rule. The predicted Meyer-Neldel temperatures are consistent with observations in several materials. The computed field-dependence of conductivity agrees with experimental data. The present work suggests that the upper temperature limit of variable range hopping is proportional to the frequency of first peak of phonon spectrum. We have also improved the description of the photocurrent decay at low temperatures. Analysis of the transition probability from an extended state to a localized state suggests that there exists a short-lifetime belt of extended states inside the conduction band or valence band.

cond-mat.mes-hall↗

arXiv2009

Phonon driven transport in amorphous semiconductors: Transition probabilities

Inspired by Holstein's work on small polaron hopping, the evolution equations of localized states and extended states in presence of atomic vibrations are derived for an amorphous semiconductor. The transition probabilities are obtained for four types of transitions: from one localized state to another localized state, from a localized state to an extended state, from an extended state to a localized state, and from one extended state to another extended state. At a temperature not too low, any process involving localized state is activated. The computed mobility of the transitions between localized states agrees with the observed `hopping mobility'. We suggest that the observed `drift mobility' originates from the transitions from localized states to extended states. Analysis of the transition probability from an extended state to a localized state suggests that there exists a short-lifetime belt of extended states inside conduction band or valence band. It agrees with the fact that photoluminescence lifetime decreases with frequency in a-Si/SiO$_{2}$ quantum well while photoluminescence lifetime is not sensitive to frequency in c-Si/SiO$_{2}$ structure.

cond-mat.mtrl-sci↗

arXiv2009

An extension of the Kubo-Greenwood formula for use in molecular simulations

We discuss the foundations and extend the range of applicability of the widely used Kubo-Greenwood formula (KGF) for the electronic conductivity. The conductivity is derived from the current density, and only the probability amplitude rather than the transition probability is used. It is shown that the contribution to the conductivity from degenerate states in a low or zero frequency external electric field and the contribution from states near resonance with a finite frequency external field are finite. The improved conductivity expression does not include the familiar "energy conserving" delta function, and no artificial broadening parameter for delta function is required for the DC limit. We explored two methods of computing current density. We discuss the role of many-electron statistics in computing the conductivity in single-particle approximations, and we show that the conventional KGF is due to the contribution from single-particle excited states.

cond-mat.stat-mech↗