Diverse and Effective Red Teaming with Auto-generated Rewards and Multi-step Reinforcement Learning

• URL: http://arxiv.org/abs/2412.18693v1
• Date: Tue, 24 Dec 2024 22:38:46 GMT
• Title: Diverse and Effective Red Teaming with Auto-generated Rewards and Multi-step Reinforcement Learning
• Authors: Alex Beutel, Kai Xiao, Johannes Heidecke, Lilian Weng,
• Abstract summary: We provide methods that enable automated red teaming to generate a large number of diverse and successful attacks. Our approach decomposes the task into two steps: (1) automated methods for generating diverse attack goals and (2) generating effective attacks for those goals.
• Score: 7.670515615413488
• License:
• Abstract: Automated red teaming can discover rare model failures and generate challenging examples that can be used for training or evaluation. However, a core challenge in automated red teaming is ensuring that the attacks are both diverse and effective. Prior methods typically succeed in optimizing either for diversity or for effectiveness, but rarely both. In this paper, we provide methods that enable automated red teaming to generate a large number of diverse and successful attacks. Our approach decomposes the task into two steps: (1) automated methods for generating diverse attack goals and (2) generating effective attacks for those goals. While we provide multiple straightforward methods for generating diverse goals, our key contributions are to train an RL attacker that both follows those goals and generates diverse attacks for those goals. First, we demonstrate that it is easy to use a large language model (LLM) to generate diverse attacker goals with per-goal prompts and rewards, including rule-based rewards (RBRs) to grade whether the attacks are successful for the particular goal. Second, we demonstrate how training the attacker model with multi-step RL, where the model is rewarded for generating attacks that are different from past attempts further increases diversity while remaining effective. We use our approach to generate both prompt injection attacks and prompts that elicit unsafe responses. In both cases, we find that our approach is able to generate highly-effective and considerably more diverse attacks than past general red-teaming approaches.

Related papers

• Target-driven Attack for Large Language Models [14.784132523066567]
  We propose our target-driven black-box attack method to maximize the KL divergence between the conditional probabilities of clean text and the attack text. Experimental results on multiple Large Language Models and datasets demonstrate the effectiveness of our attack method.
  arXiv  Detail & Related papers  (2024-11-09T15:59:59Z)
• Derail Yourself: Multi-turn LLM Jailbreak Attack through Self-discovered Clues [88.96201324719205]
  This study exposes the safety vulnerabilities of Large Language Models (LLMs) in multi-turn interactions. We introduce ActorAttack, a novel multi-turn attack method inspired by actor-network theory.
  arXiv  Detail & Related papers  (2024-10-14T16:41:49Z)
• Holistic Automated Red Teaming for Large Language Models through Top-Down Test Case Generation and Multi-turn Interaction [24.499874512829198]
  We proposeHolistic Automated Red teaMing, which scales up the diversity of test cases based on an adversarial, fine-grained risk taxonomy. Our method also leverages a novel fine-tuning strategy and reinforcement learning techniques to facilitate multi-turn probing in a human-like manner.
  arXiv  Detail & Related papers  (2024-09-25T09:44:48Z)
• DiveR-CT: Diversity-enhanced Red Teaming Large Language Model Assistants with Relaxing Constraints [68.82294911302579]
  We introduce DiveR-CT, which relaxes conventional constraints on the objective and semantic reward, granting greater freedom for the policy to enhance diversity. Our experiments demonstrate DiveR-CT's marked superiority over baselines by 1) generating data that perform better in various diversity metrics across different attack success rate levels, 2) better-enhancing resiliency in blue team models through safety tuning based on collected data, 3) allowing dynamic control of objective weights for reliable and controllable attack success rates, and 4) reducing susceptibility to reward overoptimization.
  arXiv  Detail & Related papers  (2024-05-29T12:12:09Z)
• 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)
• Multi-granular Adversarial Attacks against Black-box Neural Ranking Models [111.58315434849047]
  We create high-quality adversarial examples by incorporating multi-granular perturbations. We transform the multi-granular attack into a sequential decision-making process. Our attack method surpasses prevailing baselines in both attack effectiveness and imperceptibility.
  arXiv  Detail & Related papers  (2024-04-02T02:08:29Z)
• LAS-AT: Adversarial Training with Learnable Attack Strategy [82.88724890186094]
  "Learnable attack strategy", dubbed LAS-AT, learns to automatically produce attack strategies to improve the model robustness. Our framework is composed of a target network that uses AEs for training to improve robustness and a strategy network that produces attack strategies to control the AE generation.
  arXiv  Detail & Related papers  (2022-03-13T10:21:26Z)
• MUTEN: Boosting Gradient-Based Adversarial Attacks via Mutant-Based Ensembles [16.424441015545252]
  MUTEN is a low-cost method to improve the success rate of well-known attacks against gradient-masking models. We show that MUTEN can increase the success rate of four attacks by up to 0.45.
  arXiv  Detail & Related papers  (2021-09-27T07:15:01Z)
• Automatic Curriculum Learning through Value Disagreement [95.19299356298876]
  Continually solving new, unsolved tasks is the key to learning diverse behaviors. In the multi-task domain, where an agent needs to reach multiple goals, the choice of training goals can largely affect sample efficiency. We propose setting up an automatic curriculum for goals that the agent needs to solve. We evaluate our method across 13 multi-goal robotic tasks and 5 navigation tasks, and demonstrate performance gains over current state-of-the-art methods.
  arXiv  Detail & Related papers  (2020-06-17T03:58:25Z)

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.