Mansi Phute

Transferrable Visual Input based Steering for Output Redirection in Large Vision Language Models

A Unified Evaluation System for Configurable Attacks in Differentiable Environments

Bridging Differentiable Rendering and Photorealistic Simulation for End-to-end Adversarial Attacks

Visual Input based Steering for Output Redirection in Large Vision Language Models

Downstream models are vulnerable to blackbox attacks by 3dgs camouflages

A Large-Scale Dataset for Testing Robustness of Image Classifiers

Interactive Visual Attribution for LLM Generation

By Self Examination, LLMs Know They Are Being Tricked!

Architectural Design Principles for Adversarially Robust CNNs

Survey on Adversarial Attacks Using Differentiable Rendering

Survey on LLM interpretation and safety