Research & Teaching Assitant, M.Sc. Computational Linguistics (IMS),Universität Stuttgart
Oct. 2025 - March 2026

• Explainable Audio Deepfake Detection — Contributing to research targeting the ESDD Challenge and Interspeech Conference

  • Analyzed silence/speech token distributions in self-supervised speech models (XLSR, HuBERT) to identify interpretable evidence for deepfake artifacts

  • Investigated neuron-level activation patterns; identified property neurons and group neurons relevant to authenticity signals

  • Applied data augmentation via neural audio codecs and vocoders (HiFi-GAN, EnCodec, BigVGAN, Stable Audio VAE) using the Self-Supervised Speech Pretrained Representation Learning (S3PRL).

• Teaching Assistant — Introduction to Deep Learning for Speech and Language Processing
  for 1st semester M.Sc. students | NLP and deep learning fundamentals

Research Assitant, ISW, Universität Stuttgart
Jul. 2025 - Sep. 2025

• Model Context Protocol (MCP) — Proof of Concept

  • Designed and tested a custom MCP server to enable agentic AI workflows in the ISW

  • Evaluated for LLM tool-use scenarios

Research Engineer, Research Center for Information Technology & Innovation, Academia Sinica
Nov. 2020 - March. 2024

• Published Research — Co-authored 5 peer-reviewed papers at these venues (ACL 2023, EACL 2023, ICAIF 2024, TREC 2021–2022) covering NLP, information retrieval, and graph-based finance AI.
  See Publications section.

• Industry-Partnered Projects

  • E.SUN Commercial Bank — CTR Prediction — Implemented Deep & Cross Network v2 with Mixture-of-Experts (MoE), achieving a 19% precision improvement over the gradient boosting model (LightGBM) baseline

  • MetaEdge Corporation — Chinese Legal Retrieval — Built a hybrid BM25 + pre-trained LM pipeline for Chinese legal document retrieval, delivered as a production-ready service

• Cheng-Wei Lin*, Yu-Pao Tu, Chuan-Ju Wang. (2024, Nov). Time-aware Graph Attention Networks for Multiperiod Default Prediction. In Proceedings of the 5th ACM International Conference on AI in Finance (ICAIF '24). Association for Computing Machinery, Brooklyn, NY, USA. https://doi.org/10.1145/3677052.3698619. 

• Jia-Huei Ju, Yu-Shiang Huang, Cheng-Wei Lin, Che Lin, Chuan-Ju Wang. (2023, July). A Compare-and-contrast Multistage Pipeline for Uncovering Financial Signals in Financial Reports. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (1, 14307--14321). Association for Computational Linguistics. https://doi.org/10.18653/v1/2023.acl-long.800. 

• Ta-Wei Huang, Jia-Huei Ju, Yu-Shiang Huang, Cheng-Wei Lin, Yi-Shyuan Chiang, Chuan-Ju Wang. (2023, May). FISH: A Financial Interactive System for Signal Highlighting. In Proceedings of the 17th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations (pp. 50--56). Association for Computational Linguistics. https://aclanthology.org/2023.eacl-demo.7. 

• Jia-Huei Ju, Wei-Chih Chen, Heng-Ta Chang, Cheng-Wei Lin, Ming-Feng Tsai, Chuan-Ju Wang. (2022, November). CFDA & CLIP at TREC 2022 NeuCLIR Track. In Proceedings of the 31st Text REtrieval Conference, TREC (Vol.22). https://trec.nist.gov/pubs/trec31/papers/CFDA_CLIP.N.pdf. 

• Jia-Huei Ju, Chih-Ting Yeh, Cheng-Wei Lin, Chia-Ying Tsao, Jun-En Ding, Ming-Feng Tsai, Chuan-Ju Wang. (2021, November). An Exploration Study of Multi-stage Conversational Passage Retrieval: Paraphrase Query Expansion and Multi-view Point-wise Ranking. In Proceedings of the 30th Text REtrieval Conference, TREC (Vol.21). https://trec.nist.gov/pubs/trec30/papers/CFDA_CLIP-CAsT.pdf. 

• 中文 母語 Chinese Native 你好，我是程緯，很高興認識你！
  臺語 母語 Taiwanese Native 我嘛會曉講臺語 / Hokkien！

• English C1｜TOEFL iBT: 103 / 120｜Reading: 24．Listening: 28．Speaking: 24．Writing: 27

• Deutsch B1｜Mündliche: 1.6．Schriftliche: 2.0｜Ich komme aus Taiwan und möchte mein Deutsch weiter verbessern!

• 日本語 N3｜大学で2年間、日本語を学びました。よろしくお願いします！

• Hobbies: I am now training Hyrox. That's my strava. I like to see quantifiable progress!
  Cooking and eating healthy diet are my own way of relaxing. I like to cook and share dishes with people as an act of service. Also exploring all types of cuisines. That's my insta posts only for food. Recently also been trying some creative art works like pottery, calligraphy, guitar. 

• Community and connection: I spent two years in summer volunteering at Orchid Island to support in their communities, I like to know different cultures, how life can be lived differently. I was supporting international student reception at NTU too, and working at NETivism that provides IT services for NPOs / NGOs across Taiwan. Knowing that technologies can be served directly to the civil society is something important for me. These experiences had shaped my believe in technologies should be served for people, not replacing people. So I prefer IA - intelligence augmentation than AI.  

• Living and working:
  (0-18yr): Grew up in Tainan, southern part of Taiwan
  (18-23yr): Studying in NTU, Taipei, the capital of Taiwan.
  (21yr): First time abroad in the Philipines (PADI Open Water Diver).
  (21yr): Internship in the US for 3 months
  (22-23yr): Exchange in Singapore for a year and then Covid happened
  (23yr): Served in military
  (23-27yr): Working in Academia Sinica, Taipei, Taiwan. 
  (28yr-Now): Studying Master, Stuttgart Germany