sciwork 2026

Chun-Shih Chang

Chun-Shih Chang works in Acadamic Sininca. He is interested in machine learning, high performance computing and C++ software development.

  • Beyond Calling BLAS: Implementing and Optimizing Matrix Multiplication
Dboy Liao

目前任職於 MaiCoin 集團,主要工作內容集中在交易策略開發,也是 bbgo 主要貢獻者之一。
興趣是讓寫在紙上的算式可以跑在電腦裡,所以除了交易外也接觸過以下領域:
- large scale data engineering
- machine learning
- deep learning
所以玩過一些像是 Spark, Kafka, Tensorflow, PyTorch 之類的,
也都歡迎來找我閒聊這些東西。
大概就是個樣樣通樣樣鬆的軟體工程師 ^_<

  • Shall We Test? BBGO 可測試交易策略開發
Junli Xiao
  • 打開 AI 的黑盒子:機制可解釋性(Mechanistic Interpretability)的最新進展
Kevin

目前就讀大學三年級,兼具混音工程師、音樂製作人與軟體工程師背景,也是 Echo Canvas 開發者與 AI 人機互動研究者。從聲音製作的實務出發,探索音訊技術、互動介面與 AI 協作如何結合,讓創作想法轉化為可操作、可驗證的工具。

在 OpenAI Build Week 期間,與 Codex 協作開發 Echo Canvas,將自然語言場景描述、空間編輯與確定性聲學計算整合為瀏覽器空間音訊原型。關注開發者如何向 AI 表達意圖、檢查生成結果,並結合工程測試與實際聆聽評估使用體驗。希望透過跨越音樂製作、軟體工程與人機互動的實作經驗,與社群交流 AI 輔助創作及開發的可能性與限制。

  • 從自然語言到空間音訊:用 Codex 打造可驗證的瀏覽器聲學原型
Kito Cheng

Open Source Compiler Developer from SiFive

  • AI Agent x Open Source Compiler
Sheng-Wen Cheng

Sheng-Wen (Colin) Cheng is a roboticist and software engineer who builds low-level software at the intersection of robotics, control engineering, and real-time systems.

He is also an active open-source contributor and has presented his work at leading open-source events, including Open Source Summit organized by the Linux Foundation.

For more information, please visit shengwen-tw.github.io

  • On the Tenth Anniversary of AlphaGo: An Introduction to Reinforcement Learning and Planning
Sivan
  • Werilog: Toward a Bidirectional Coding Architecture
Tetsuo Koyama

I am an active contributor to PyVista—a VTK-based Python library for 3D visualization—regularly submitting pull requests for bug fixes and feature enhancements. Expanding on this work, I joined the pyvista-wasm project to bring PyVista to WebAssembly, where I now serve as one of the lead developers driving its architecture and implementation. Although I have been involved for less than a year, I have spearheaded core technical efforts, including establishing the build pipeline using Pyodide and VTK-WASM and integrating in-browser rendering. All of this work is done independently as an open-source contributor in my personal time.

  • PyVista on WebAssembly: Enabling Serverless 3D Visualization
Timmy

軟體工程師,專長於雲端系統架構與全端開發。近期致力於研究 LLM 底層原理,希望對日常工作中頻繁使用的 AI 模型有更深入的理解。

  • 從 MicroGPT 到真實模型:拆解 LLM 原理,再用 LoRA 微調打造 Python 文件繁中翻譯 LLM
William Mou

熱衷於太空與天文領域的軟體工程師。因 NASA Hackathon 與夥伴結緣,曾在清大參與衛星計畫與天文觀測,近期更投入火箭導控軟體競賽開發。不過,我的本職其實是 HPC 及 RISC-V 軟體效能工程師。

  • 體驗火箭導控 Balloon Popping Challenge