相关论文: When Prohibitions Become Permissions: Auditing Neg…

As artificial intelligence (AI) systems play an increasingly prominent role in human decision-making, challenges surface in the realm of human-AI interactions. One challenge arises from the suboptimal AI policies due to the inadequate…

Most frameworks for assessing the openness of AI systems use narrow criteria such as availability of data, model, code, documentation, and licensing terms. However, to evaluate whether the intended effects of openness - such as…

An auditor instructs an AI assistant: "open each file individually using the Read tool -- no scripts, no agents." The AI replies "Yes" -- then issues a single batched call summarizing all fifty files at once. We call this the Compliance…

When AI systems explain their reasoning step-by-step, practitioners often assume these explanations reveal what actually influenced the AI's answer. We tested this assumption by embedding hints into questions and measuring whether models…

We examine two properties of AI systems: capability (what a system can do) and steerability (how reliably one can shift behavior toward intended outcomes). A central question is whether capability growth reduces steerability and risks…

Current alignment evaluation mostly measures whether models encode dangerous concepts and whether they refuse harmful requests. Both miss the layer where alignment often operates: routing from concept detection to behavioral policy. We…

Previous research on expert advice-taking shows that humans exhibit two contradictory behaviors: on the one hand, people tend to overvalue their own opinions undervaluing the expert opinion, and on the other, people often defer to other…

Large language models are trained to refuse harmful requests, but can they accurately predict when they will refuse before responding? We investigate this question through a systematic study where models first predict their refusal…

As artificial intelligence systems become increasingly agentic, capable of general reasoning, planning, and value prioritization, current safety practices that treat obedience as a proxy for ethical behavior are becoming inadequate. This…

In decision support applications of AI, the AI algorithm's output is framed as a suggestion to a human user. The user may ignore this advice or take it into consideration to modify their decision. With the increasing prevalence of such…

Since the release of OpenAI's ChatGPT, generative language models have attracted extensive public attention. The increased usage has highlighted generative models' broad utility, but also revealed several forms of embedded bias. Some is…

Scalable oversight protocols aim to empower evaluators to accurately verify AI models more capable than themselves. However, human evaluators are subject to biases that can lead to systematic errors. We conduct two studies examining the…

Safety-aligned language models often refuse prompts that are actually harmless. Current evaluations mostly report global rates such as false rejection or compliance. These scores treat each prompt alone and miss local inconsistency, where a…

As AI agents attempt to autonomously act on users' behalf, they raise transparency and control issues. We argue that permission-based access control is indispensable in providing meaningful control to the users, but conventional permission…

If AI models can detect when they are being evaluated, the effectiveness of evaluations might be compromised. For example, models could have systematically different behavior during evaluations, leading to less reliable benchmarks for…

Negation is a common linguistic phenomenon. Yet language models face challenges with negation in many natural language understanding tasks such as question answering and natural language inference. In this paper, we experiment with seamless…

Agentic systems increasingly rely on language models to monitor their own behavior. For example, coding agents may self critique generated code for pull request approval or assess the safety of tool-use actions. We show that this design…

Negation instructions such as 'do not mention $X$' can paradoxically increase the accessibility of $X$ in human thought, a phenomenon known as ironic rebound. Large language models (LLMs) face the same challenge: suppressing a concept…

A key component of building safe and reliable language models is enabling the models to appropriately refuse to follow certain instructions or answer certain questions. We may want models to output refusal messages for various categories of…

AI-assisted code review tools typically operate as generic "expert reviewer" agents, producing homogeneous findings regardless of the analysis type needed. We present a system that constrains AI reviewer behavior through philosophical…