Related papers: Jailbreaking as a Reward Misspecification Problem
Recent studies have discovered that large language models (LLM) may be ``fooled'' to output private information, including training data, system prompts, and personally identifiable information, under carefully crafted adversarial prompts.…
Large language models have drawn significant attention to the challenge of safe alignment, especially regarding jailbreak attacks that circumvent security measures to produce harmful content. To address the limitations of existing methods…
"Jailbreak" is a major safety concern of Large Language Models (LLMs), which occurs when malicious prompts lead LLMs to produce harmful outputs, raising issues about the reliability and safety of LLMs. Therefore, an effective evaluation of…
Red-teaming is a common practice for mitigating unsafe behaviors in Large Language Models (LLMs), which involves thoroughly assessing LLMs to identify potential flaws and addressing them with responsible and accurate responses. While…
The widespread deployment of large language models (LLMs) has raised growing concerns about their misuse risks and associated safety issues. While prior studies have examined the safety of LLMs in general usage, code generation, and…
Pluralistic alignment has emerged as a critical frontier in the development of Large Language Models (LLMs), with reward models (RMs) serving as a central mechanism for capturing diverse human values. While benchmarks for general response…
This paper proposes a jailbreaking prompt detection method for large language models (LLMs) to defend against jailbreak attacks. Although recent LLMs are equipped with built-in safeguards, it remains possible to craft jailbreaking prompts…
Large language models (LLMs) increasingly operate in multi-agent and safety-critical settings, raising open questions about how their vulnerabilities scale when models interact adversarially. This study examines whether larger models can…
Recently, Large Language Models (LLMs) have made significant advancements and are now widely used across various domains. Unfortunately, there has been a rising concern that LLMs can be misused to generate harmful or malicious content.…
A fundamental issue in deep learning has been adversarial robustness. As these systems have scaled, such issues have persisted. Currently, large language models (LLMs) with billions of parameters suffer from adversarial attacks just like…
Jailbreak vulnerabilities in Large Language Models (LLMs) refer to methods that extract malicious content from the model by carefully crafting prompts or suffixes, which has garnered significant attention from the research community.…
Reward modeling (RM), which captures human preferences to align large language models (LLMs), is increasingly employed in tasks such as model finetuning, response filtering, and ranking. However, due to the inherent complexity of human…
Large language models~(LLMs) are expected to be helpful, harmless, and honest. In different alignment scenarios, such as safety, confidence, and general preference alignment, binary preference data collection and reward modeling are…
Alignment in large language models (LLMs) is used to enforce guidelines such as safety. Yet, alignment fails in the face of jailbreak attacks that modify inputs to induce unsafe outputs. In this paper, we introduce and evaluate a new…
Reinforcement learning from human feedback (RLHF) has emerged as a powerful technique to make large language models (LLMs) more capable in complex settings. RLHF proceeds as collecting human preference data, training a reward model on said…
Jailbreak attacks represent one of the most sophisticated threats to the security of large language models (LLMs). To deal with such risks, we introduce an innovative framework that can help evaluate the effectiveness of jailbreak attacks…
We identify a structural weakness in current large language model (LLM) alignment: modern refusal mechanisms are fail-open. While existing approaches encode refusal behaviors across multiple latent features, suppressing a single dominant…
Large language models (LLMs) generate human-aligned content under certain safety constraints. However, the current known technique ``jailbreak prompt'' can circumvent safety-aligned measures and induce LLMs to output malicious content.…
Reinforcement Learning (RL) algorithms for safety alignment of Large Language Models (LLMs), such as Direct Preference Optimization (DPO), encounter the challenge of distribution shift. Current approaches typically address this issue…
The rapid adoption of large language models (LLMs) in financial services introduces new operational, regulatory, and security risks. Yet most red-teaming benchmarks remain domain-agnostic and fail to capture failure modes specific to…