System-Technology Co-Evaluation of A7 CFET and A10 NSFET Technologies from Cell Parasitics to Chip Reliability
本文通过物理基础的热和老化感知系统技术协同评估方法，解决了A7 CFET与A10 NSFET技术中从单元寄生到芯片可靠性的问题。
0 citationsRead paper
Selected work
Representative Papers
Evaluating LLM-based AI agents integrated with materials synthesis tools: the case of atomic layer deposition
本文探讨了基于大语言模型的AI代理在材料合成中的性能评估策略，以原子层沉积为例，提出了一种实用的评估框架。
0 citationsRead paper
RecipeNet: A Hierarchical Transformer for Recipe Data
This study addresses the limitations of existing tabular learning methods in capturing hierarchical field interactions and process dependencies within recipe data. To overcome these challenges, this work proposes RecipeNet, which employs a hierarchical Transformer architecture. By stacking encoders to jointly model intra-step field interactions and inter-step sequential dependencies, the method effectively integrates both structured and sequential characteristics of recipes. Experimental results demonstrate that RecipeNet consistently outperforms state-of-the-art tabular models across multiple benchmark datasets. These findings validate the efficacy of hierarchical sequence modeling for complex recipe representation learning and establish a novel paradigm for industrial recipe data mining.
0 citationsRead paper
Reinforcement Learning for Autonomous Warehouse Orchestration in SAP Logistics Execution: Redefining Supply Chain Agility
In SAP Logistics Execution (LE) systems, warehouse task coordination suffers from delayed response times, while conventional rule engines struggle to adapt to dynamic disruptions and multilingual operational contexts. Method: This paper proposes the first end-to-end reinforcement learning (RL)-based autonomous orchestration framework tailored for SAP LE. It integrates Deep Q-Networks (DQN) with state abstraction modeling, trained on a synthetically generated dataset of 300,000 SAP LE transaction records—each annotated with multilingual metadata and diverse exception events. A lightweight API layer enables high-fidelity, privacy-preserving system integration while ensuring industrial scalability. Contribution/Results: The framework achieves 95% task optimization accuracy and reduces average processing time by 60%. It supports real-time heatmap visualization and closed-loop agile decision-making, effectively overcoming the responsiveness limitations of traditional rule-based engines in dynamic, heterogeneous warehouse environments.
0 citationsRead paper
Recent publications
Latest Papers
System-Technology Co-Evaluation of A7 CFET and A10 NSFET Technologies from Cell Parasitics to Chip Reliability
本文通过物理基础的热和老化感知系统技术协同评估方法，解决了A7 CFET与A10 NSFET技术中从单元寄生到芯片可靠性的问题。
0 citationsRead paper
Evaluating LLM-based AI agents integrated with materials synthesis tools: the case of atomic layer deposition
本文探讨了基于大语言模型的AI代理在材料合成中的性能评估策略，以原子层沉积为例，提出了一种实用的评估框架。
0 citationsRead paper
RecipeNet: A Hierarchical Transformer for Recipe Data
This study addresses the limitations of existing tabular learning methods in capturing hierarchical field interactions and process dependencies within recipe data. To overcome these challenges, this work proposes RecipeNet, which employs a hierarchical Transformer architecture. By stacking encoders to jointly model intra-step field interactions and inter-step sequential dependencies, the method effectively integrates both structured and sequential characteristics of recipes. Experimental results demonstrate that RecipeNet consistently outperforms state-of-the-art tabular models across multiple benchmark datasets. These findings validate the efficacy of hierarchical sequence modeling for complex recipe representation learning and establish a novel paradigm for industrial recipe data mining.
0 citationsRead paper
Reinforcement Learning for Autonomous Warehouse Orchestration in SAP Logistics Execution: Redefining Supply Chain Agility
In SAP Logistics Execution (LE) systems, warehouse task coordination suffers from delayed response times, while conventional rule engines struggle to adapt to dynamic disruptions and multilingual operational contexts. Method: This paper proposes the first end-to-end reinforcement learning (RL)-based autonomous orchestration framework tailored for SAP LE. It integrates Deep Q-Networks (DQN) with state abstraction modeling, trained on a synthetically generated dataset of 300,000 SAP LE transaction records—each annotated with multilingual metadata and diverse exception events. A lightweight API layer enables high-fidelity, privacy-preserving system integration while ensuring industrial scalability. Contribution/Results: The framework achieves 95% task optimization accuracy and reduces average processing time by 60%. It supports real-time heatmap visualization and closed-loop agile decision-making, effectively overcoming the responsiveness limitations of traditional rule-based engines in dynamic, heterogeneous warehouse environments.
0 citationsRead paper