Benyamin Tafreshian

|

AI Security Researcher & Threat Hunter

Focusing on adversarial machine learning, network intrusion detection, LLM security, and threat detection. Working on building practical solutions using AI to improve security, robustness, and comprehensibility of complex systems.

Terminal

Terminal initialized...

Type 'help' or press Tab for available commands.

Projects

96-bit Block Cipher

Custom cryptographic block cipher implementation with 96-bit key

C++CMakePython

ParamAtlas

Zero-knob inspector for LLMs/VLMs

PythonPyTorchTransformers

RedditEngage ML Predictor

Machine learning model for predicting Reddit post engagement

SparkMLlibScikit-learn

NeuroViz Neural Network Visualization

Interactive visualization tool for neural network architectures

TypeScriptReactNext.js

Fraud Detection Big Data Pipeline

Big data pipeline for real-time fraud detection

SparkKafkaCassandra

Publications

RoguePrompt: Dual-Layer Ciphering for Self-Reconstruction to Circumvent LLM Moderation

Benyamin Tafreshian, Prathamesh Dhake

2026 • arXiv preprint

Abstract PDF

A Defensive Framework Against Adversarial Attacks on Machine Learning-Based Network Intrusion Detection Systems

Benyamin Tafreshian, Shengzhi Zhang

2024 • IEEE 23rd International Conference on Trust, Security and Privacy in Computing and Communications (TrustCom 2024)

DOI Publisher Abstract PDF

Certifications

CompTIA Security+

VERIFIED

CEH (Ethical Hacker)

ACTIVE

CISSP Associate

PENDING

OSCP

IN_PROGRESS

Contact

Mail GitHub LinkedIn Scholar

Documents

Resume Academic CV

System Info

Fingerprint

Visitor IDLoading...

Canvas FP—

Audio FP—

System

OS— —

Browser— —

Engine—

Device—

Brand—

Model—

Fonts0

Plugins0

Color Gamut—

HDRNO

Hardware

CPU Cores8

Architecture—

RAM8 GB

Resolution1920×1080

Color Depth24-bit

TouchNo

GPU Vendor—

GPU—

GPU Perf—

Privacy

Ad Blockers—

CookiesYes

Local StorageYES

PDF ViewerNO

DNTEnabled (Do Not Track)

Network

LocationLoading..., Loading...

TimezoneEurope/Berlin

Languageen-US