About

Dr. Wen Chen is an Associate Professor of Aerospace and Mechanical Engineering and Materials Science at the University of Southern California (USC). He obtained his Ph.D. in Mechanical Engineering and Materials Science from Yale University in 2016. His research interests include additive manufacturing (3D printing), architected materials, and physical and mechanical metallurgy. Dr. Chen's work focuses on understanding the fundamental processing-structure-property relationships in advanced materials and integrating control over materials on multiple length scales through additive manufacturing to achieve extreme properties. Prior to joining USC in 2025, he served as an Associate Professor in Mechanical and Industrial Engineering at the University of Massachusetts Amherst and worked as a postdoctoral research scientist at Lawrence Livermore National Laboratory, where he studied various additive manufacturing techniques such as projection stereolithography, direct ink writing, laser powder-bed fusion, and laser-directed energy deposition. He has published over 100 papers in refereed journals, holds 5 US patents, and has served as an associate editor for journals including Materials Futures and Scientific Reports. His research activities include developing additive manufacturing technologies, high throughput materials design, and designing and fabricating architected materials for structural and energy applications. Dr. Chen has received numerous awards, including the NSF CAREER Award in 2023, the TMS Young Leaders Professional Development Award in 2025, and the Barbara H. and Joseph J. Goldstein Outstanding Junior Faculty Award in 2024.

Research topics

• Computer Science
• Computer network
• Telecommunications
• Computer Security
• Artificial Intelligence
• Engineering
• Operating system
• Algorithm
• Electronic engineering
• World Wide Web
• Mathematical optimization
• Physics

Selected publications

• Safe Delta: Consistently Preserving Safety when Fine-Tuning LLMs on Diverse Datasets

  ArXiv.org · 2025-05-17

  preprintOpen access

  Large language models (LLMs) have shown great potential as general-purpose AI assistants across various domains. To fully leverage this potential in specific applications, many companies provide fine-tuning API services, enabling users to upload their own data for LLM customization. However, fine-tuning services introduce a new safety threat: user-uploaded data, whether harmful or benign, can break the model's alignment, leading to unsafe outputs. Moreover, existing defense methods struggle to address the diversity of fine-tuning datasets (e.g., varying sizes, tasks), often sacrificing utility for safety or vice versa. To address this issue, we propose Safe Delta, a safety-aware post-training defense method that adjusts the delta parameters (i.e., the parameter change before and after fine-tuning). Specifically, Safe Delta estimates the safety degradation, selects delta parameters to maximize utility while limiting overall safety loss, and applies a safety compensation vector to mitigate residual safety loss. Through extensive experiments on four diverse datasets with varying settings, our approach consistently preserves safety while ensuring that the utility gain from benign datasets remains unaffected.

  PublisherOA PDFDOI

• Extraction and Characterization of Sustainable Cellulosic Fibers from <i>Urera Hypselodendron</i> Plant

  Journal of Natural Fibers · 2025-07-09 · 2 citations

  articleOpen access1st author

  Nowadays, researchers are paying more attention to natural fibers because they are under-dominated by synthetic fibers. This is implied in terms of their low cost, renewability, biodegradability, and the need for low energy for their production. The purpose of this study is to extract bast fiber from the plant using water retting and alkali treatment. Analysis of the composition of fibers extracted by water retting revealed 71% cellulose, 15% hemicellulose, and 17% lignin, while 74% cellulose, 11% hemicellulose, and 12.6% lignin by 3% NaOH. Fiber diameter was measured at 106 μm for water-retted fibers and 97 μm for alkali-extracted fibers. The tensile strength was 4.2 cN/dtex for water-retted fibers, and 5.6 cN/dtex for alkali-extracted fibers. The extracted fiber analysis using Fourier transform infrared spectroscopy (FT-IR) was used to evaluate the functional groups, and the chemical bonds on the fiber molecule. The moisture content of the fiber was 7.14% by water retting and 6.8% by 3% NaOH less than the jute and flax fibers. The fineness of 1.87 tex water retting decreased from 1.02 tex. Fibers could be used as an alternative material for composites and other technical textile applications.

  PublisherDOI

• Plasma-treated conductive textile advancements in coating and functional properties: A review

  Materials Today Sustainability · 2025-12-13

  reviewOpen accessSenior author

  Despite their significant contribution to wearable electronic applications, conductive textiles face practical performance limitations due to the intrinsically insulating nature of textile fibers and the poor durability, adhesion, and low conductivity of traditional conductive polymer coatings. Materials like PEDOT: PSS, polypyrrole, graphene, and metal nanoparticles, all of which coat fibrous substrates non-uniformly, resulting in poor charge transport and high contact resistance. Unfortunately, these failures lead to rapid degradation in terms of either shortening the service life of electrical performance under mechanical deformation, washing, or long-term use. It limits their integration in reliable sensors, energy-harvesting devices, and health monitoring systems. This review demonstrates how cold plasma techniques are used to address such persistent drawbacks. Plasma-induced functional groups enhance the surface energy and introduce nanoscale roughness to provide strong adhesion interface with coatings while producing improved interfacial bonding. Thus, conductive polymers, MXenes, and metal-polymer nanocomposite coatings through plasma-assisted deposition exhibit comparatively less electrical resistance with superior mechanical properties, retaining the flexibility and breathability of the fabric. Additionally, the plasma-enabled coatings confer multifunctional properties such as antibacterial, photothermal, and stable bio signals in sensing. The review finally identifies future challenges-enhanced scalability, long-term electrical stability under extreme conditions, and a sustainable process-while highlighting emerging opportunities associated with plasma-engineered textiles for next-generation smart wearables.

  PublisherDOI

• Advances in nanocrystalline cellulosic materials: Biorefinery perspectives and biomedical innovations; A comprehensive review

  Materials Today Chemistry · 2025-07-01

  reviewSenior authorCorresponding
  PublisherDOI

• A Large Language Model-based Multi-Agent Framework for Analog Circuits’ Sizing Relationships Extraction

  2025-05-09 · 2 citations

  article

  In the design process of the analog circuit pre-layout phase, device sizing is an important step in determining whether an analog circuit can meet the required performance metrics. Many existing techniques extract the circuit sizing task as a mathematical optimization problem to solve and continuously improve the optimization efficiency from a mathematical perspective. But they ignore the automatic introduction of prior knowledge, fail to achieve effective pruning of the search space, which thereby leads to a considerable compression margin remaining in the search space. To alleviate this problem, we propose a large language model (LLM)-based multi-agent framework for analog circuits’ sizing relationships extraction from academic papers. The search space in the sizing process can be effectively pruned based on the sizing relationship extracted by this framework. Eventually, we conducted tests on 3 types of circuits, and the optimization efficiency was improved by 2.32~26×. work demonstrates that the LLM can effectively prune the search space for analog circuit sizing, providing a new solution for the combination of LLMs and conventional analog circuit design automation methods.

  PublisherDOI

• Cobalt atom incorporation regulates grain structure and thermoelectric properties in Bi₂Te₃ thin films

  Surface and Coatings Technology · 2025-09-22

  article
  PublisherDOI

• A solar energy system with a dual-input power converter and global MPPT for off-grid applications

  Electric Power Systems Research · 2025-02-06 · 4 citations

  article
  PublisherDOI

• Optimization of Cotton Fabric Dyeing Using <i>Urera Hypselodendron</i> Leaf Extract and Wood Ash as a Bio-Mordant

  Journal of Natural Fibers · 2025-08-25

  articleOpen accessSenior authorCorresponding

  In recent years, natural dyes have gained importance in textile dyeing applications, as many synthetic dyes have been found to be carcinogenic, mutagenic, and allergenic. This study investigates the use of natural dyes derived from Urera hypselodendron leaves, in combination with wood ash as a bio-mordant, for dyeing cotton fabrics. Aluminum sulfate (5%) and iron (II) sulfate (1%) were used as control mordants for comparison with the ash-based mordant. For the dyeing process, three common mordanting techniques were applied: pre-mordanting, meta-mordanting, and post-mordanting. Pre-mordanting with aluminum sulfate resulted in the highest color strength and overall fastness properties. Using the pre-mordanting method, the ash-mordanted samples produced a dull yellowish color with uniform dye uptake, achieving a maximum color strength (K/S value) of 2.704. Process optimization was carried out by varying the dyeing parameters such as time (30–75 min), temperature (45–105°C), and dye concentration (2–6% o.w.f.). The optimal dyeing conditions included a mordant concentration of 4.99%, a dyeing time of 51.59 min, a dye amount of 37.75 mL, and a dyeing temperature of 72.25°C. These results aim to develop a sustainable and eco-friendly dyeing method suitable for textile industries seeking alternatives to conventional chemical-based processes.

  PublisherOA PDFDOI

• AmpAgent: An LLM-based Multi-Agent System for Multi-stage Amplifier Schematic Design from Literature for Process and Performance Porting

  arXiv (Cornell University) · 2024-09-23 · 8 citations

  preprintOpen access

  Multi-stage amplifiers are widely applied in analog circuits. However, their large number of components, complex transfer functions, and intricate pole-zero distributions necessitate extensive manpower for derivation and param sizing to ensure their stability. In order to achieve efficient derivation of the transfer function and simplify the difficulty of circuit design, we propose AmpAgent: a multi-agent system based on large language models (LLMs) for efficiently designing such complex amplifiers from literature with process and performance porting. AmpAgent is composed of three agents: Literature Analysis Agent, Mathematics Reasoning Agent and Device Sizing Agent. They are separately responsible for retrieving key information (e.g. formulas and transfer functions) from the literature, decompose the whole circuit's design problem by deriving the key formulas, and address the decomposed problem iteratively. AmpAgent was employed in the schematic design of seven types of multi-stage amplifiers with different compensation techniques. In terms of design efficiency, AmpAgent has reduced the number of iterations by 1.32$ \sim $4${\times}$ and execution time by 1.19$ \sim $2.99${\times}$ compared to conventional optimization algorithms, with a success rate increased by 1.03$ \sim $6.79${\times}$. In terms of circuit performance, it has improved by 1.63$ \sim $27.25${\times}$ compared to the original literature. The findings suggest that LLMs could play a crucial role in the field of complex analog circuit schematic design, as well as process and performance porting.

  PublisherOA PDFDOI

• Exploring the Development Path of Female Leading Characters in Film and Television Works - Taking the Image Shaping of Female Characters as an Example

  Lecture Notes in Education Psychology and Public Media · 2024-07-31

  articleOpen access1st authorCorresponding

  In recent years, people have been paying more and more attention to the development direction of the "women" group, and the image of women in society has become increasingly diverse. In today's society, film and television media are important channels that can quickly disseminate information, attract social attention, and disseminate value orientation. This paper explores the development path of female protagonists by examining the image shaping of female characters in film and television works. By analyzing the evolution of female characters in various types of film and television productions, one can see that female roles are no longer confined to traditional archetypes such as the virtuous wife and loving mother or tragic figures. Instead, they increasingly exhibit qualities of independence, bravery, confidence, and intelligence. These changes enrich the content of film and television works and provide audiences with more diverse values and role models. The study finds that the diversification of female character portrayals not only enhances the artistic quality and appeal of the works but also exerts a positive influence on society. The image of women in film and television is gradually becoming an important force in guiding social trends and promoting gender equality. Therefore, in-depth exploration and research into the development path of female characters in film and television works are of great significance for understanding social and cultural changes and promoting gender equality.

  PublisherOA PDFDOI

Frequent coauthors

• M.A. Breuer

  12 shared

• Ren‐Hung Hwang

  11 shared

• Wen-Tsuen Chen

  National Tsing Hua University

  10 shared

• Masaaki Ito

  KDDI Research (Japan)

  9 shared

• Sandeep K. Gupta

  University of Southern California

  9 shared

• Jenq‐Gong Duh

  8 shared

• Chih-Yu Wang

  Research Center for Information Technology Innovation, Academia Sinica

  7 shared

• Bor‐Sen Chen

  Yuan Ze University

  7 shared

Education

• Master, Electrical Engineering

  National Tsing Hua University

  2016

Awards & honors

• TMS Young Leaders Professional Development Award (2025)
• Barbara H. and Joseph J. Goldstein Outstanding Junior Facult…
• 35 Emerging Young Investigators Under 35(ish) highlighted by…
• NSF CAREER Award (2023)
• SME Outstanding Young Manufacturing Engineer (2022)