Finding trojans in aligned LLMs. Official repository for the competition hosted at SaTML 2024.
☆119Jun 13, 2024Updated 2 years ago
Alternatives and similar repositories for rlhf_trojan_competition
Users that are interested in rlhf_trojan_competition are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Code for paper "Universal Jailbreak Backdoors from Poisoned Human Feedback"☆67Apr 24, 2024Updated 2 years ago
- PAL: Proxy-Guided Black-Box Attack on Large Language Models☆57Aug 17, 2024Updated last year
- ☆19Feb 25, 2024Updated 2 years ago
- This is the starter kit for the Trojan Detection Challenge 2023 (LLM Edition), a NeurIPS 2023 competition.☆92May 19, 2024Updated 2 years ago
- Adversarial Attacks on GPT-4 via Simple Random Search [Dec 2023]☆43Apr 28, 2024Updated 2 years ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Improving Alignment and Robustness with Circuit Breakers☆266Sep 24, 2024Updated last year
- On the Loss Landscape of Adversarial Training: Identifying Challenges and How to Overcome Them [NeurIPS 2020]☆36Jul 3, 2021Updated 5 years ago
- This is the official Gtihub repo for our paper: "BEEAR: Embedding-based Adversarial Removal of Safety Backdoors in Instruction-tuned Lang…☆23Jul 3, 2024Updated 2 years ago
- Easily deploy my zsh and tmux configuration on new machines. Includes local and remote aliases to improve workflow.☆15Apr 23, 2026Updated 3 months ago
- Improved Few-Shot Jailbreaking Can Circumvent Aligned Language Models and Their Defenses (NeurIPS 2024)☆65Jan 11, 2025Updated last year
- Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks [ICLR 2025]☆391Jan 23, 2025Updated last year
- A collection of different ways to implement accessing and modifying internal model activations for LLMs☆24Oct 18, 2024Updated last year
- ☆35Sep 13, 2023Updated 2 years ago
- ☆61Mar 9, 2023Updated 3 years ago
- Open source password manager - Proton Pass • AdSecurely store, share, and autofill your credentials with Proton Pass, the end-to-end encrypted password manager trusted by millions.
- ☆15Jul 12, 2024Updated 2 years ago
- ☆18Aug 15, 2022Updated 3 years ago
- ☆31Nov 11, 2025Updated 8 months ago
- ☆30Jun 19, 2023Updated 3 years ago
- [ICLR 2025] Cheating Automatic LLM Benchmarks: Null Models Achieve High Win Rates (Oral)☆86Oct 23, 2024Updated last year
- Code for the paper "A Light Recipe to Train Robust Vision Transformers" [SaTML 2023]☆54Feb 6, 2023Updated 3 years ago
- Code for T-MARS data filtering☆35Aug 23, 2023Updated 2 years ago
- ☆20Feb 11, 2024Updated 2 years ago
- ICLR2024 Paper. Showing properties of safety tuning and exaggerated safety.☆95May 9, 2024Updated 2 years ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- Code for "Prior Convictions: Black-Box Adversarial Attacks with Bandits and Priors"☆65Jan 14, 2020Updated 6 years ago
- ☆71Feb 4, 2024Updated 2 years ago
- [arXiv:2311.03191] "DeepInception: Hypnotize Large Language Model to Be Jailbreaker"☆177Feb 20, 2024Updated 2 years ago
- Contains random samples referenced in the paper "Sleeper Agents: Training Robustly Deceptive LLMs that Persist Through Safety Training".☆150Mar 9, 2024Updated 2 years ago
- Codes for NeurIPS 2021 paper "Adversarial Neuron Pruning Purifies Backdoored Deep Models"☆65May 8, 2023Updated 3 years ago
- HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal☆1,018Aug 16, 2024Updated last year
- Repo for the research paper "SecAlign: Defending Against Prompt Injection with Preference Optimization"☆98Jul 2, 2026Updated 3 weeks ago
- Code for the paper "Evading Black-box Classifiers Without Breaking Eggs" [SaTML 2024]☆21Apr 15, 2024Updated 2 years ago
- ☆17Sep 4, 2024Updated last year
- Deploy on Railway without the complexity - Free Credits Offer • AdConnect your repo and Railway handles the rest with instant previews. Quickly provision container image services, databases, and storage volumes.
- ☆15Jul 24, 2022Updated 4 years ago
- A benchmark for mechanistic discovery of circuits in Transformers☆17Dec 15, 2024Updated last year
- ☆203Nov 26, 2023Updated 2 years ago
- ☆24Jul 25, 2024Updated 2 years ago
- Representation Engineering: A Top-Down Approach to AI Transparency☆1,015Aug 14, 2024Updated last year
- [ICCV 2023 Oral] Official implementation of "Robust Evaluation of Diffusion-Based Adversarial Purification"☆26Dec 3, 2023Updated 2 years ago
- Implementation of our ICLR 2021 paper: Policy-Driven Attack: Learning to Query for Hard-label Black-box Adversarial Examples.☆11Mar 9, 2021Updated 5 years ago