microsoft / TaskTrackerLinks
TaskTracker is an approach to detecting task drift in Large Language Models (LLMs) by analysing their internal activations. It provides a simple linear probe-based method and a more sophisticated metric learning method to achieve this. The project also releases the computationally expensive activation data to stimulate further AI safety research…
☆69Updated last month
Alternatives and similar repositories for TaskTracker
Users that are interested in TaskTracker are comparing it to the libraries listed below
Sorting:
- A benchmark for evaluating the robustness of LLMs and defenses to indirect prompt injection attacks.☆87Updated last year
- A Dynamic Environment to Evaluate Attacks and Defenses for LLM Agents.☆328Updated 2 weeks ago
- [NDSS'25 Best Technical Poster] A collection of automated evaluators for assessing jailbreak attempts.☆172Updated 6 months ago
- Repo for the research paper "SecAlign: Defending Against Prompt Injection with Preference Optimization"☆72Updated 3 months ago
- A repository of Language Model Vulnerabilities and Exposures (LVEs).☆112Updated last year
- PAL: Proxy-Guided Black-Box Attack on Large Language Models☆55Updated last year
- ☆85Updated last year
- This repository provides a benchmark for prompt injection attacks and defenses☆310Updated 2 weeks ago
- The official implementation of our pre-print paper "Automatic and Universal Prompt Injection Attacks against Large Language Models".☆60Updated last year
- TAP: An automated jailbreaking method for black-box LLMs☆194Updated 10 months ago
- ☆65Updated 10 months ago
- ☆49Updated last year
- PromptInject is a framework that assembles prompts in a modular fashion to provide a quantitative analysis of the robustness of LLMs to a…☆427Updated last year
- Code&Data for the paper "Watch Out for Your Agents! Investigating Backdoor Threats to LLM-Based Agents" [NeurIPS 2024]☆93Updated last year
- Code to generate NeuralExecs (prompt injection for LLMs)☆25Updated 3 weeks ago
- Implementation of BEAST adversarial attack for language models (ICML 2024)☆91Updated last year
- ☆111Updated last year
- Package to optimize Adversarial Attacks against (Large) Language Models with Varied Objectives☆69Updated last year
- Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks [ICLR 2025]☆357Updated 9 months ago
- ☆94Updated 11 months ago
- ☆35Updated 11 months ago
- Finding trojans in aligned LLMs. Official repository for the competition hosted at SaTML 2024.☆115Updated last year
- [NeurIPS 2024] Official implementation for "AgentPoison: Red-teaming LLM Agents via Memory or Knowledge Base Backdoor Poisoning"☆160Updated 6 months ago
- Open One-Stop Moderation Tools for Safety Risks, Jailbreaks, and Refusals of LLMs☆92Updated 10 months ago
- [ICML 2024] COLD-Attack: Jailbreaking LLMs with Stealthiness and Controllability☆166Updated 10 months ago
- Fluent student-teacher redteaming☆23Updated last year
- official implementation of [USENIX Sec'25] StruQ: Defending Against Prompt Injection with Structured Queries☆48Updated 3 months ago
- Code for the paper "Defeating Prompt Injections by Design"☆138Updated 4 months ago
- A fast + lightweight implementation of the GCG algorithm in PyTorch☆293Updated 5 months ago
- Papers about red teaming LLMs and Multimodal models.☆145Updated 5 months ago