An open source implementation of CLIP.

An interactive NVIDIA-GPU process viewer and beyond, the one-stop solution for GPU process management.

[ICML'24] SeeAct is a system for generalist web agents that autonomously carry out tasks on any given website, with a focus on large multimodal models (LMMs) such as GPT-4V(ision).

The official implementation of our ICLR2024 paper "AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language Models".

An Open Robustness Benchmark for Jailbreaking Language Models [arXiv 2024]

[ECCV 2024] The official code for "AdaShield: Safeguarding Multimodal Large Language Models from Structure-based Attack via Adaptive Shield Prompting."

JailBreakV-28K: A comprehensive benchmark designed to evaluate the transferability of LLM jailbreak attacks to MLLMs, and further assess the robustness and safety of MLLMs against a variety of jail…

A lightweight library for large laguage model (LLM) jailbreaking defense.

The official implementation of our pre-print paper "Automatic and Universal Prompt Injection Attacks against Large Language Models".

Data and Code for the paper, Knowledge-to-Jailbreak: One Knowledge Point Worth One Attack.