Penghui Yang

Penghui Yang 杨鹏辉

Email: phyang [dot] cs [at] gmail [dot] com

Penghui Yang is currently a PhD candidate at the College of Computing and Data Science, Nanyang Technological University, supervised by Prof. Bo An. He received his B.Sc. degree in Computer Science from Nanjing University of Aeronautics and Astronautics in 2023, advised by Prof. Sheng-Jun Huang. Previously, he collaborated closely with Dr. Ming-Kun Xie and Prof. Lei Feng. He also spent a cherished year and a half at Sea AI Lab, working under the mentorship of Dr. Cunxiao Du.

My research focuses on knowledge distillation, speculative decoding, and AI for Science through software-in-the-loop systems. Across these areas, I study how paired interactions between complementary models and computational tools make learning, inference, and scientific discovery more efficient.

LearningKnowledge Distillation

Teacher

structured supervision↓

Student

Design supervision that transfers useful knowledge to compact models, improving capability under deployment constraints.

MLKD · DHKD

InferenceSpeculative Decoding

Draft model

propose↓verify↑

Target model

Combine inexpensive proposals with target verification to accelerate generation without changing the target distribution.

LongSpec · Sailor2

Software-in-the-LoopAI for Science

Neural agent

propose↓feedback↑

Scientific software

Close the loop between neural reasoning and physics-based computation to refine designs, recover calculations, and investigate mechanisms.

AutoMAT · AutoDFT · XDFT

Penghui Yang 杨鹏辉

News

Research Highlights

LearningKnowledge Distillation

InferenceSpeculative Decoding

Software-in-the-LoopAI for Science

Publications ( show selected / show all by date / show all by topic )

Discovering Physical Mechanisms from Experiment-Simulation Mismatches

Yue Li, Penghui Yang, Yushan Xiao, Zhonghan Zhang, Jianguo Huang, Yuhao Lu, Cuntai Guan, Bo An, Bijun Tang, Zheng Liu

FedDCA: Robust Federated Learning via Fine-Grained Channel Anomaly Detection and Consensus-Constrained Aggregation

Binbin Ding, Penghui Yang, Shuxian Xiong, Pei Peng, Zeqing Ge, Sheng-Jun Huang^†

MATAI: A Unified Interactive Platform for AI-Driven Alloy Discovery

Yixuan Li^, Yanchen Deng^, Chendong Zhao^*, Penghui Yang, Jianguo Huang, Yuhao Lu, Yushan Xiao, Xuyu Dong, Zhonghan Zhang, Cuntai Guan, Bijun Tang^†, Xinrun Wang^†, Zheng Liu^†, Bo An^†

D³AS: Dynamic Deadline-Driven Architecture Search for Chronologically Challenged Researchers

Penghui Yang

AutoDFT: A Closed-Loop Multi-Agent Framework for Autonomous DFT Calculations

Penghui Yang^, Zhonghan Zhang^, Yue Li^*, Xinrun Wang^†, Yanchen Deng, Yuhao Lu, Bijun Tang^†, Zheng Liu, Bo An^†

LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification

Penghui Yang^†, Cunxiao Du^†, Fengzhuo Zhang^†, Haonan Wang, Tianyu Pang, Chao Du, Bo An

MatSeek: An Automated Knowledge-Driven Framework for Materials Research

Jianguo Huang^, Xuyu Dong^, Yuhao Lu, Yanchen Deng, Chendong Zhao, Penghui Yang, Zhonghan Zhang, Yixuan Li, Yushan Xiao, Cuntai Guan^†, Bijun Tang^†, Xinrun Wang^†, Zheng Liu^†, Bo An^†

Anatomy of a Hybrid Mind: Deconstructing Hybrid Reasoning in Large Language Models

Penghui Yang^, Fengzhuo Zhang^, Yunlong Hou, Tianyu Pang, Chao Du, Bo An, Cunxiao Du^†

MATAI: A Generalist Machine Learning Framework for Property Prediction and Inverse Design of Advanced Alloys

Yanchen Deng^, Chendong Zhao^, Yixuan Li^*, Bijun Tang^†, Xinrun Wang^†, Zhonghan Zhang, Yuhao Lu, Penghui Yang, Jianguo Huang, Yushan Xiao, Cuntai Guan^†, Zheng Liu^†, Bo An^†

Autonomous Multi-objective Alloy Design through Simulation-guided Optimization

Penghui Yang^, Chendong Zhao^, Bijun Tang^†, Zhonghan Zhang, Xinrun Wang^†, Yanchen Deng, Xuyu Dong, Yuhao Lu, Jianguo Huang, Yixuan Li, Yushan Xiao, Cuntai Guan, Zheng Liu^†, Bo An^†

Dual-Head Knowledge Distillation: Enhancing Logits Utilization with an Auxiliary Head

Penghui Yang, Chen-Chen Zong, Sheng-Jun Huang, Lei Feng^†, Bo An

FedDLAD: A Federated Learning Dual-Layer Anomaly Detection Framework for Enhancing Resilience Against Backdoor Attacks

Binbin Ding, Penghui Yang, Sheng-Jun Huang^†

FLAIN: Mitigating Backdoor Attacks in Federated Learning via Flipping Weight Updates of Low-Activation Input Neurons

Binbin Ding, Penghui Yang, Sheng-Jun Huang^†

A Unified Open Adapter for Open-World Noisy Label Learning: Data-Centric and Learning-Based Insights

Chen-Chen Zong, Penghui Yang, Ming-Kun Xie, Sheng-Jun Huang^†

Sailor2: Sailing in South-East Asia with Inclusive Multilingual LLMs

Sailor2 Team

Multi-Label Knowledge Distillation

Penghui Yang^, Ming-Kun Xie^, Chen-Chen Zong, Lei Feng, Gang Niu, Masashi Sugiyama, Sheng-Jun Huang^†

Penghui Yang 杨鹏辉

News

Research Highlights

LearningKnowledge Distillation

InferenceSpeculative Decoding

Software-in-the-LoopAI for Science

Publications ( show selected / show all by date / show all by topic )

Discovering Physical Mechanisms from Experiment-Simulation Mismatches

Yue Li, Penghui Yang, Yushan Xiao, Zhonghan Zhang, Jianguo Huang, Yuhao Lu, Cuntai Guan, Bo An, Bijun Tang, Zheng Liu

FedDCA: Robust Federated Learning via Fine-Grained Channel Anomaly Detection and Consensus-Constrained Aggregation

Binbin Ding, Penghui Yang, Shuxian Xiong, Pei Peng, Zeqing Ge, Sheng-Jun Huang†

MATAI: A Unified Interactive Platform for AI-Driven Alloy Discovery

Yixuan Li*, Yanchen Deng*, Chendong Zhao*, Penghui Yang, Jianguo Huang, Yuhao Lu, Yushan Xiao, Xuyu Dong, Zhonghan Zhang, Cuntai Guan, Bijun Tang†, Xinrun Wang†, Zheng Liu†, Bo An†

D3AS: Dynamic Deadline-Driven Architecture Search for Chronologically Challenged Researchers

Penghui Yang

AutoDFT: A Closed-Loop Multi-Agent Framework for Autonomous DFT Calculations

Penghui Yang*, Zhonghan Zhang*, Yue Li*, Xinrun Wang†, Yanchen Deng, Yuhao Lu, Bijun Tang†, Zheng Liu, Bo An†

LongSpec: Long-Context Lossless Speculative Decoding with Efficient Drafting and Verification

Penghui Yang*†, Cunxiao Du*†, Fengzhuo Zhang†, Haonan Wang, Tianyu Pang, Chao Du, Bo An

MatSeek: An Automated Knowledge-Driven Framework for Materials Research

Jianguo Huang*, Xuyu Dong*, Yuhao Lu, Yanchen Deng, Chendong Zhao, Penghui Yang, Zhonghan Zhang, Yixuan Li, Yushan Xiao, Cuntai Guan†, Bijun Tang†, Xinrun Wang†, Zheng Liu†, Bo An†

Anatomy of a Hybrid Mind: Deconstructing Hybrid Reasoning in Large Language Models

Penghui Yang*, Fengzhuo Zhang*, Yunlong Hou, Tianyu Pang, Chao Du, Bo An, Cunxiao Du†

MATAI: A Generalist Machine Learning Framework for Property Prediction and Inverse Design of Advanced Alloys

Yanchen Deng*, Chendong Zhao*, Yixuan Li*, Bijun Tang†, Xinrun Wang†, Zhonghan Zhang, Yuhao Lu, Penghui Yang, Jianguo Huang, Yushan Xiao, Cuntai Guan†, Zheng Liu†, Bo An†

Autonomous Multi-objective Alloy Design through Simulation-guided Optimization

Penghui Yang*, Chendong Zhao*, Bijun Tang†, Zhonghan Zhang, Xinrun Wang†, Yanchen Deng, Xuyu Dong, Yuhao Lu, Jianguo Huang, Yixuan Li, Yushan Xiao, Cuntai Guan, Zheng Liu†, Bo An†

Dual-Head Knowledge Distillation: Enhancing Logits Utilization with an Auxiliary Head

Penghui Yang, Chen-Chen Zong, Sheng-Jun Huang, Lei Feng†, Bo An

FedDLAD: A Federated Learning Dual-Layer Anomaly Detection Framework for Enhancing Resilience Against Backdoor Attacks

Binbin Ding, Penghui Yang, Sheng-Jun Huang†

FLAIN: Mitigating Backdoor Attacks in Federated Learning via Flipping Weight Updates of Low-Activation Input Neurons

Binbin Ding, Penghui Yang, Sheng-Jun Huang†

A Unified Open Adapter for Open-World Noisy Label Learning: Data-Centric and Learning-Based Insights

Chen-Chen Zong, Penghui Yang, Ming-Kun Xie, Sheng-Jun Huang†

Sailor2: Sailing in South-East Asia with Inclusive Multilingual LLMs

Sailor2 Team

Multi-Label Knowledge Distillation

Penghui Yang*, Ming-Kun Xie*, Chen-Chen Zong, Lei Feng, Gang Niu, Masashi Sugiyama, Sheng-Jun Huang†

Binbin Ding, Penghui Yang, Shuxian Xiong, Pei Peng, Zeqing Ge, Sheng-Jun Huang^†

Yixuan Li^, Yanchen Deng^, Chendong Zhao^*, Penghui Yang, Jianguo Huang, Yuhao Lu, Yushan Xiao, Xuyu Dong, Zhonghan Zhang, Cuntai Guan, Bijun Tang^†, Xinrun Wang^†, Zheng Liu^†, Bo An^†

D³AS: Dynamic Deadline-Driven Architecture Search for Chronologically Challenged Researchers

Penghui Yang^, Zhonghan Zhang^, Yue Li^*, Xinrun Wang^†, Yanchen Deng, Yuhao Lu, Bijun Tang^†, Zheng Liu, Bo An^†

Penghui Yang^†, Cunxiao Du^†, Fengzhuo Zhang^†, Haonan Wang, Tianyu Pang, Chao Du, Bo An

Jianguo Huang^, Xuyu Dong^, Yuhao Lu, Yanchen Deng, Chendong Zhao, Penghui Yang, Zhonghan Zhang, Yixuan Li, Yushan Xiao, Cuntai Guan^†, Bijun Tang^†, Xinrun Wang^†, Zheng Liu^†, Bo An^†

Penghui Yang^, Fengzhuo Zhang^, Yunlong Hou, Tianyu Pang, Chao Du, Bo An, Cunxiao Du^†

Yanchen Deng^, Chendong Zhao^, Yixuan Li^*, Bijun Tang^†, Xinrun Wang^†, Zhonghan Zhang, Yuhao Lu, Penghui Yang, Jianguo Huang, Yushan Xiao, Cuntai Guan^†, Zheng Liu^†, Bo An^†

Penghui Yang^, Chendong Zhao^, Bijun Tang^†, Zhonghan Zhang, Xinrun Wang^†, Yanchen Deng, Xuyu Dong, Yuhao Lu, Jianguo Huang, Yixuan Li, Yushan Xiao, Cuntai Guan, Zheng Liu^†, Bo An^†

Penghui Yang, Chen-Chen Zong, Sheng-Jun Huang, Lei Feng^†, Bo An

Binbin Ding, Penghui Yang, Sheng-Jun Huang^†

Binbin Ding, Penghui Yang, Sheng-Jun Huang^†

Chen-Chen Zong, Penghui Yang, Ming-Kun Xie, Sheng-Jun Huang^†

Penghui Yang^, Ming-Kun Xie^, Chen-Chen Zong, Lei Feng, Gang Niu, Masashi Sugiyama, Sheng-Jun Huang^†