A research workbench for developing and testing attacks against large language models, with a focus on prompt injection vulnerabilities and defenses.

dInfer: An Efficient Inference Framework for Diffusion Language Models

Flexible and Pluggable Serving Engine for Diffusion LLMs

Gemma open-weight LLM library, from Google DeepMind

The best ChatGPT that $100 can buy.

[EMNLP 2025 Oral] IPIGuard: A Novel Tool Dependency Graph-Based Defense Against Indirect Prompt Injection in LLM Agents

Curated resources, research, and tools for securing AI systems

gpt-oss-120b and gpt-oss-20b are two open-weight language models by OpenAI

Renderer for the harmony response format to be used with gpt-oss

Repo for the paper "Meta SecAlign: A Secure Foundation LLM Against Prompt Injection Attacks".

Code for the paper "Defeating Prompt Injections by Design"

[ICLR 2025] Dissecting adversarial robustness of multimodal language model agents

Open-source implementation of AlphaEvolve

Official PyTorch implementation for "Large Language Diffusion Models"

OO for LLMs

Dataset and code for "JailbreaksOverTime: Detecting Jailbreak Attacks Under Distribution Shift"

Bag of Tricks: Benchmarking of Jailbreak Attacks on LLMs. Empirical tricks for LLM Jailbreaking. (NeurIPS 2024)

Repo for the research paper "SecAlign: Defending Against Prompt Injection with Preference Optimization"

A Dynamic Environment to Evaluate Attacks and Defenses for LLM Agents.

A reading list for large models safety, security, and privacy (including Awesome LLM Security, Safety, etc.).

A Survey on Jailbreak Attacks and Defenses against Multimodal Generative Models

A data augmentations library for audio, image, text, and video.

Fast near-duplicate matching is a method for quickly finding near-duplicate spans in a document by utilizing the Rabin-Karp algorithm.

The Security Toolkit for LLM Interactions

LLM Prompt Injection Detector

Every practical and proposed defense against prompt injection.