Skip to content
View keyanUB's full-sized avatar
🏠
Working from home
🏠
Working from home
8 followers · 4 following

Achievements

Achievement: QuickdrawAchievement: Pull Shark

Achievements

Achievement: QuickdrawAchievement: Pull Shark

Highlights

  • Pro

Organizations

@CactiLab

Block or report keyanUB

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.

Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. 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
keyanUB/README.md

Hi, I'm Keyan

Ph.D. candidate at University at Buffalo · AI security · Generative AI safety

Typing SVG

View repositories Personal website LinkedIn

Daily Reflection

A daily thought with its source and bilingual translation.

Today's Thought · 2026-07-29

What you do today can improve all your tomorrows.
Ralph Marston

你今天做的事，可以改善所有的明天。

Open positive quote card

Dashboard

Current Focus

  • Generative AI security and safety
  • Adversarial robustness for multimodal systems
  • Online abuse and harmful content moderation

Tech Stack

Tech stack icons

Pinned Loading

  1. cuadvancelab.github.io cuadvancelab.github.io Public

    Forked from cuadvancelab/cuadvancelab.github.io

    cu advance lab website

    Jupyter Notebook

  2. CactiLab/HateGuard CactiLab/HateGuard Public

    Code for S&P 2024 paper: Moderating New Waves of Online Hate with Chain-of-Thought Reasoning in Large Language Models.

    Jupyter Notebook 11 2

  3. UBSec/UGCG-Guard UBSec/UGCG-Guard Public

    Code for USENIX Security 2024 paper: Moderating Illicit Online Image Promotion for Unsafe User Generated Content Games Using Large Vision-Language Models.

    Python 13

  4. Puliczek/awesome-mcp-security Puliczek/awesome-mcp-security Public

    🔥🔒 Awesome MCP (Model Context Protocol) Security 🖥️

    724 183

  5. Prompt-Optimization-Agent Prompt-Optimization-Agent Public

    Auto-optimize your LLM prompts with a multi-step AI pipeline: complexity scoring, analysis, rewriting, and execution. Built for personal use, open to all.

    Python