FLRT: Fluent Student-Teacher Redteaming

• URL: http://arxiv.org/abs/2407.17447v2
• Date: Tue, 1 Oct 2024 17:39:09 GMT
• Title: FLRT: Fluent Student-Teacher Redteaming
• Authors: T. Ben Thompson, Michael Sklar,
• Abstract summary: We improve existing algorithms to develop powerful and fluent attacks on safety-tuned models. Our technique centers around a new distillation-based approach that encourages the victim model to emulate a toxified finetune. We achieve attack success rates $>93$% for Llama-2-7B, Llama-3-8B, and Vicuna-7B, while maintaining model-measured perplexity $33$.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Many publicly available language models have been safety tuned to reduce the likelihood of toxic or liability-inducing text. To redteam or jailbreak these models for compliance with toxic requests, users and security analysts have developed adversarial prompting techniques. One attack method is to apply discrete optimization techniques to the prompt. However, the resulting attack strings are often gibberish text, easily filtered by defenders due to high measured perplexity, and may fail for unseen tasks and/or well-tuned models. In this work, we improve existing algorithms (primarily GCG and BEAST) to develop powerful and fluent attacks on safety-tuned models like Llama-2 and Phi-3. Our technique centers around a new distillation-based approach that encourages the victim model to emulate a toxified finetune, either in terms of output probabilities or internal activations. To encourage human-fluent attacks, we add a multi-model perplexity penalty and a repetition penalty to the objective. We also enhance optimizer strength by allowing token insertions, token swaps, and token deletions and by using longer attack sequences. The resulting process is able to reliably jailbreak the most difficult target models with prompts that appear similar to human-written prompts. On Advbench we achieve attack success rates $>93$% for Llama-2-7B, Llama-3-8B, and Vicuna-7B, while maintaining model-measured perplexity $<33$; we achieve $95$% attack success for Phi-3, though with higher perplexity. We also find a universally-optimized single fluent prompt that induces $>88$% compliance on previously unseen tasks across Llama-2-7B, Phi-3-mini and Vicuna-7B and transfers to other black-box models.

Related papers

• A Realistic Threat Model for Large Language Model Jailbreaks [87.64278063236847]
  In this work, we propose a unified threat model for the principled comparison of jailbreak attacks. Our threat model combines constraints in perplexity, measuring how far a jailbreak deviates from natural text. We adapt popular attacks to this new, realistic threat model, with which we, for the first time, benchmark these attacks on equal footing.
  arXiv  Detail & Related papers  (2024-10-21T17:27:01Z)
• Deciphering the Chaos: Enhancing Jailbreak Attacks via Adversarial Prompt Translation [71.92055093709924]
  We propose a novel method that "translates" garbled adversarial prompts into coherent and human-readable natural language adversarial prompts. It also offers a new approach to discovering effective designs for jailbreak prompts, advancing the understanding of jailbreak attacks. Our method achieves over 90% attack success rates against Llama-2-Chat models on AdvBench, despite their outstanding resistance to jailbreak attacks.
  arXiv  Detail & Related papers  (2024-10-15T06:31:04Z)
• Improved Generation of Adversarial Examples Against Safety-aligned LLMs [72.38072942860309]
  Adversarial prompts generated using gradient-based methods exhibit outstanding performance in performing automatic jailbreak attacks against safety-aligned LLMs. In this paper, we explore a new perspective on this problem, suggesting that it can be alleviated by leveraging innovations inspired in transfer-based attacks. We show that 87% of the query-specific adversarial suffixes generated by the developed combination can induce Llama-2-7B-Chat to produce the output that exactly matches the target string on AdvBench.
  arXiv  Detail & Related papers  (2024-05-28T06:10:12Z)
• Revisiting Character-level Adversarial Attacks for Language Models [53.446619686108754]
  We introduce Charmer, an efficient query-based adversarial attack capable of achieving high attack success rate (ASR) Our method successfully targets both small (BERT) and large (Llama 2) models.
  arXiv  Detail & Related papers  (2024-05-07T14:23:22Z)
• Goal-guided Generative Prompt Injection Attack on Large Language Models [6.175969971471705]
  Large language models (LLMs) provide a strong foundation for large-scale user-oriented natural language tasks. A large number of users can easily inject adversarial text or instructions through the user interface. It is unclear how these strategies relate to the success rate of attacks and thus effectively improve model security.
  arXiv  Detail & Related papers  (2024-04-06T06:17:10Z)
• Jailbreaking Leading Safety-Aligned LLMs with Simple Adaptive Attacks [38.25697806663553]
  We show that even the most recent safety-aligned LLMs are not robust to simple adaptive jailbreaking attacks. We achieve 100% attack success rate -- according to GPT-4 as a judge -- on Vicuna-13B, Mistral-7B, Phi-3-Mini, Nemotron-4-340B, Llama-2-Chat-7B/13B/70B, Llama-3-Instruct-8B, Gemma-7B, GPT-3.5, GPT-4o, and R2D2 from HarmBench.
  arXiv  Detail & Related papers  (2024-04-02T17:58:27Z)
• AutoDAN: Interpretable Gradient-Based Adversarial Attacks on Large Language Models [55.748851471119906]
  Safety alignment of Large Language Models (LLMs) can be compromised with manual jailbreak attacks and (automatic) adversarial attacks. Recent studies suggest that defending against these attacks is possible: adversarial attacks generate unlimited but unreadable gibberish prompts, detectable by perplexity-based filters. We introduce AutoDAN, an interpretable, gradient-based adversarial attack that merges the strengths of both attack types.
  arXiv  Detail & Related papers  (2023-10-23T17:46:07Z)
• Learning to Ignore Adversarial Attacks [14.24585085013907]
  We introduce the use of rationale models that can explicitly learn to ignore attack tokens. We find that the rationale models can successfully ignore over 90% of attack tokens.
  arXiv  Detail & Related papers  (2022-05-23T18:01:30Z)
• Multi-granularity Textual Adversarial Attack with Behavior Cloning [4.727534308759158]
  We propose MAYA, a Multi-grAnularitY Attack model to generate high-quality adversarial samples with fewer queries to victim models. We conduct comprehensive experiments to evaluate our attack models by attacking BiLSTM, BERT and RoBERTa in two different black-box attack settings and three benchmark datasets.
  arXiv  Detail & Related papers  (2021-09-09T15:46:45Z)
• Towards Variable-Length Textual Adversarial Attacks [68.27995111870712]
  It is non-trivial to conduct textual adversarial attacks on natural language processing tasks due to the discreteness of data. In this paper, we propose variable-length textual adversarial attacks(VL-Attack) Our method can achieve $33.18$ BLEU score on IWSLT14 German-English translation, achieving an improvement of $1.47$ over the baseline model.
  arXiv  Detail & Related papers  (2021-04-16T14:37:27Z)

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.