Skip to content
View wangxiaoyang0412's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report wangxiaoyang0412

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
wangxiaoyang0412/README.md
Xiaoyang Wang — PhD Candidate, Drexel University. Agentic AI / Multimodal Learning / Trustworthy AI

Homepage Google Scholar LinkedIn X Email ORCID

Note

I'm on the 2027 job market (PhD expected Spring 2027), open to industry research scientist roles and academic postdoc / faculty positions in trustworthy and medical AI. Email is the best way to reach me: xw388@drexel.edu.

I'm a PhD candidate at Drexel University, advised by Prof. Christopher C. Yang. Clinical AI can be accurate on average yet still fail the patients who need it most, by being unfair, opaque or confidently wrong. My dissertation asks when multi-agent LLM systems in medicine deserve our trust: when the agents agree, do they actually reason alike?

  • Agentic AI: multi-agent LLM systems for clinical reasoning, and when their consensus can be trusted
  • Multimodal learning: robust patient models from whatever combination of modalities is available
  • Trustworthy AI: clinical prediction that is fair, explainable and knows when to abstain

Toolbox

Python, PyTorch, Hugging Face, vLLM, scikit-learn, pandas

Pinned Loading

  1. consistency-illusion consistency-illusion Public

    [EMNLP 2026 Main] The Consistency Illusion: CARA metrics for cross-agent reasoning alignment in multi-agent LLM debate, and the Grounded Debate Protocol (GDP) for medical QA

    Python 1

  2. probeguard probeguard Public

    [NeurIPS 2026 GenAI4Health] Certified abstention for multi-round medical LLM consensus: process features, rationale semantic entropy, a counter-evidence probe, and stratified Learn-then-Test calibr…

    Python

  3. disparity-laundering disparity-laundering Public

    [EACL 2027] Clinical multi-agent LLM teams launder demographic bias into consensus; the Diverse-Perspective Ensemble cuts it by up to 94% with no retraining

    Python