Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts
- URL: http://arxiv.org/abs/2402.16822v2
- Date: Mon, 22 Jul 2024 17:31:43 GMT
- Title: Rainbow Teaming: Open-Ended Generation of Diverse Adversarial Prompts
- Authors: Mikayel Samvelyan, Sharath Chandra Raparthy, Andrei Lupu, Eric Hambro, Aram H. Markosyan, Manish Bhatt, Yuning Mao, Minqi Jiang, Jack Parker-Holder, Jakob Foerster, Tim Rocktäschel, Roberta Raileanu,
- Abstract summary: We present Rainbow Teaming, a novel black-box approach for producing a diverse collection of adversarial prompts.
Our approach reveals hundreds of effective adversarial prompts, with an attack success rate exceeding 90%.
We additionally explore the versatility of Rainbow Teaming by applying it to question answering and cybersecurity.
- Score: 57.49685172971446
- License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
- Abstract: As large language models (LLMs) become increasingly prevalent across many real-world applications, understanding and enhancing their robustness to adversarial attacks is of paramount importance. Existing methods for identifying adversarial prompts tend to focus on specific domains, lack diversity, or require extensive human annotations. To address these limitations, we present Rainbow Teaming, a novel black-box approach for producing a diverse collection of adversarial prompts. Rainbow Teaming casts adversarial prompt generation as a quality-diversity problem, and uses open-ended search to generate prompts that are both effective and diverse. Focusing on the safety domain, we use Rainbow Teaming to target various state-of-the-art LLMs, including the Llama 2 and Llama 3 models. Our approach reveals hundreds of effective adversarial prompts, with an attack success rate exceeding 90% across all tested models. Furthermore, we demonstrate that fine-tuning models with synthetic data generated by the Rainbow Teaming method significantly enhances their safety without sacrificing general performance or helpfulness. We additionally explore the versatility of Rainbow Teaming by applying it to question answering and cybersecurity, showcasing its potential to drive robust open-ended self-improvement in a wide range of applications.
Related papers
- Ferret: Faster and Effective Automated Red Teaming with Reward-Based Scoring Technique [22.2168585464366]
Ferret is a novel approach that builds upon Rainbow Teaming by generating multiple adversarial prompt mutations per iteration.
Ferret improves the overall attack success rate (ASR) to 95%, which is 46% higher than Rainbow Teaming.
arXiv Detail & Related papers (2024-08-20T09:58:01Z) - Learning diverse attacks on large language models for robust red-teaming and safety tuning [126.32539952157083]
Red-teaming, or identifying prompts that elicit harmful responses, is a critical step in ensuring the safe deployment of large language models.
We show that even with explicit regularization to favor novelty and diversity, existing approaches suffer from mode collapse or fail to generate effective attacks.
We propose to use GFlowNet fine-tuning, followed by a secondary smoothing phase, to train the attacker model to generate diverse and effective attack prompts.
arXiv Detail & Related papers (2024-05-28T19:16:17Z) - White-box Multimodal Jailbreaks Against Large Vision-Language Models [61.97578116584653]
We propose a more comprehensive strategy that jointly attacks both text and image modalities to exploit a broader spectrum of vulnerability within Large Vision-Language Models.
Our attack method begins by optimizing an adversarial image prefix from random noise to generate diverse harmful responses in the absence of text input.
An adversarial text suffix is integrated and co-optimized with the adversarial image prefix to maximize the probability of eliciting affirmative responses to various harmful instructions.
arXiv Detail & Related papers (2024-05-28T07:13:30Z) - Against The Achilles' Heel: A Survey on Red Teaming for Generative Models [60.21722603260243]
Our extensive survey, which examines over 120 papers, introduces a taxonomy of fine-grained attack strategies grounded in the inherent capabilities of language models.
We have developed the "searcher" framework to unify various automatic red teaming approaches.
arXiv Detail & Related papers (2024-03-31T09:50:39Z) - An Open-World, Diverse, Cross-Spatial-Temporal Benchmark for Dynamic Wild Person Re-Identification [58.5877965612088]
Person re-identification (ReID) has made great strides thanks to the data-driven deep learning techniques.
The existing benchmark datasets lack diversity, and models trained on these data cannot generalize well to dynamic wild scenarios.
We develop a new Open-World, Diverse, Cross-Spatial-Temporal dataset named OWD with several distinct features.
arXiv Detail & Related papers (2024-03-22T11:21:51Z) - Multi-Agent Diagnostics for Robustness via Illuminated Diversity [37.38316542660311]
We present Multi-Agent Diagnostics for Robustness via Illuminated Diversity (MADRID)
MADRID generates diverse adversarial scenarios that expose strategic vulnerabilities in pre-trained multi-agent policies.
We evaluate the effectiveness of MADRID on the 11vs11 version of Google Research Football.
arXiv Detail & Related papers (2024-01-24T14:02:09Z) - Adversarial Prompt Tuning for Vision-Language Models [86.5543597406173]
Adversarial Prompt Tuning (AdvPT) is a technique to enhance the adversarial robustness of image encoders in Vision-Language Models (VLMs)
We demonstrate that AdvPT improves resistance against white-box and black-box adversarial attacks and exhibits a synergistic effect when combined with existing image-processing-based defense techniques.
arXiv Detail & Related papers (2023-11-19T07:47:43Z)
This list is automatically generated from the titles and abstracts of the papers in this site.
This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.