Cyan H.

Key Skills

Software

No software listed

Top Subject Matter

Technology – Software Development & Code QA

AI & Data Science – LLM Alignment, RLHF & Fine-tuning

Cybersecurity – Cyber Threat Intelligence & Vulnerability Detection

Top Data Types

Computer Code Programming

Image

Text

Top Task Types

RLHF

Fine-tuning

Computer Programming/Coding

Text Summarization

Function Calling

Question Answering

Classification

Bounding Box

Freelancer Overview

Yutong Huang (cid:131) +86-177-0811-0383 # [email protected] § github.com/Hyut1212 Education Sichuan University Sep. 2022 – Present Bachelor of Science in Cybersecurity (College of Cyber Science and Engineering) Chengdu, China Technical Skills Languages: Python, C++, C, HTML/CSS Frameworks & Tools: PyTorch, Git, Linux, Docker, LaTeX LLM & Post-Training Spec: Transformer Architecture, Attention Mechanisms, Pre-training & Fine-tuning Paradigms, Post-Training Systems, Alignment & Evaluation Foundation Machine Learning & Reinforcement Learning: Supervised/Unsupervised Learning, Deep Neural Networks, Policy/Value-Based RL Methods Research Experience Sichuan University Sep. 2025 – May 2026 Undergraduate Research Assistant (Advisor: Prof. Cheng Huang) Chengdu, China • First Author of the research paper: ”CTS-OD: A Cascaded Two-Stage Framework for High-Throughput Obfs4 Traffic Detection” (Accepted by Cybersecurity Journal). • Addressed feature shift and high-throughput bottlenecks in darknet traffic obfuscated by the Obfs4 protocol by designing a cascaded, high-efficiency two-stage detection pipeline, achieving an 0.99 F1-score with state-of-the-art inference speed. • Developed a high-performance retrieval and classification engine integrating label-based HDBSCAN clustering and FAISS-accelerated vector search for rapid traffic filtering, followed by a Random Forest validator. • Engineered and curated a massive, benchmark-level encrypted traffic dataset exceeding 500 GB contain

Education

Work History