BridgeMatch: Conditional Transport Bridges in Matching Matrix Space for 3D Deformable Registration
本文提出BridgeMatch方法，通过两阶段生成解算器在匹配矩阵空间中保持完整软匹配矩阵，以解决3D可变形配准中的非刚性点云对应问题。
0 citationsRead paper
Selected work
Representative Papers
Physical Law Ecology: mapping multi-mechanism ecologies as the zeroth step of data-driven scientific discovery
该研究针对多机制系统科学发现的瓶颈，提出物理定律生态学框架，通过数据确定共存独立机制数量，自动挖掘候选方程并构建参数空间中的主导权重场。
0 citationsRead paper
Unsaid, Unsafe? Implicit Security Obligations in LLM-Based RTL Code Generation
研究解决了LLM生成的RTL代码安全性问题，通过构建SECRTL-GEN基准和RTL-Obliger框架来识别并修正安全漏洞，提高代码的安全性和功能性。
0 citationsRead paper
Execution-Anchored Hallucination Calibration Reranking for Verilog Code Generation
研究解决了低资源硬件描述语言Verilog代码生成问题，提出EAHC框架，结合执行信号和推理信号以优化多候选重排序。
0 citationsRead paper
Why does AI unlock new possibilities in STEM education? A Bibliometric Analysis of Trends and Future Agenda
This study addresses the persistent challenges in STEM education—namely, insufficient personalization and difficulties in interdisciplinary integration—by conducting a bibliometric analysis of 242 publications from 2015 to 2025. Integrating knowledge graph construction, scientometrics, and text mining, the research systematically traces the evolution of AI-enhanced STEM education, revealing a paradigm shift from intelligent tutoring toward inquiry-based learning and computational thinking development. It further elucidates, for the first time, how large language model–driven “intelligent scaffolding” lowers cognitive barriers, thereby serving as a pivotal mechanism that transitions STEM education from knowledge transmission to competence cultivation. Building on these insights, the study proposes a forward-looking research agenda to guide future scholarship in this emerging domain.
0 citationsRead paper
Recent publications
Latest Papers
BridgeMatch: Conditional Transport Bridges in Matching Matrix Space for 3D Deformable Registration
本文提出BridgeMatch方法，通过两阶段生成解算器在匹配矩阵空间中保持完整软匹配矩阵，以解决3D可变形配准中的非刚性点云对应问题。
0 citationsRead paper
Physical Law Ecology: mapping multi-mechanism ecologies as the zeroth step of data-driven scientific discovery
该研究针对多机制系统科学发现的瓶颈，提出物理定律生态学框架，通过数据确定共存独立机制数量，自动挖掘候选方程并构建参数空间中的主导权重场。
0 citationsRead paper
Unsaid, Unsafe? Implicit Security Obligations in LLM-Based RTL Code Generation
研究解决了LLM生成的RTL代码安全性问题，通过构建SECRTL-GEN基准和RTL-Obliger框架来识别并修正安全漏洞，提高代码的安全性和功能性。
0 citationsRead paper
Execution-Anchored Hallucination Calibration Reranking for Verilog Code Generation
研究解决了低资源硬件描述语言Verilog代码生成问题，提出EAHC框架，结合执行信号和推理信号以优化多候选重排序。
0 citationsRead paper
Why does AI unlock new possibilities in STEM education? A Bibliometric Analysis of Trends and Future Agenda
This study addresses the persistent challenges in STEM education—namely, insufficient personalization and difficulties in interdisciplinary integration—by conducting a bibliometric analysis of 242 publications from 2015 to 2025. Integrating knowledge graph construction, scientometrics, and text mining, the research systematically traces the evolution of AI-enhanced STEM education, revealing a paradigm shift from intelligent tutoring toward inquiry-based learning and computational thinking development. It further elucidates, for the first time, how large language model–driven “intelligent scaffolding” lowers cognitive barriers, thereby serving as a pivotal mechanism that transitions STEM education from knowledge transmission to competence cultivation. Building on these insights, the study proposes a forward-looking research agenda to guide future scholarship in this emerging domain.
0 citationsRead paper