Oyster-I：超越拒绝——负责任语言模型的建设性安全对齐
Oyster-I: Beyond Refusal -- Constructive Safety Alignment for Responsible Language Models
AI总结：
针对现有安全对齐过度依赖拒绝的问题，提出建设性安全对齐（CSA）范式，通过博弈论预测、风险边界发现和可解释推理控制，在保证安全的同时引导用户，实现接近GPT-5的建设性表现和接近GPT-o1的鲁棒性。
摘要：
大型语言模型（LLMs）通常部署安全机制以防止有害内容的生成。当前大多数方法狭隘地关注恶意行为者带来的风险，常常将风险视为对抗性事件，并依赖防御性的拒绝。然而，在现实世界场景中，风险也来自非恶意用户，他们在心理困扰（例如，自残意图）下寻求帮助。在这种情况下，模型的响应会强烈影响用户的后续行动。简单的拒绝可能导致他们重复、升级或转移到不安全的平台，从而造成更糟糕的结果。我们引入了建设性安全对齐（CSA），这是一种以人为中心的范式，既能防止恶意滥用，又能积极引导脆弱用户走向安全且有帮助的结果。CSA在Oyster-I（Oy1）中实现，结合了博弈论的用户反应预测、细粒度风险边界发现和可解释的推理控制，将安全转变为建立信任的过程。Oy1在开放模型中实现了最先进的安全性，同时保持了较高的通用能力。在我们的建设性基准测试中，它表现出强大的建设性参与度，接近GPT-5，并在Strata-Sword越狱数据集上展现出无与伦比的鲁棒性，接近GPT-o1水平。通过从拒绝优先转向引导优先的安全策略，CSA重新定义了模型与用户的关系，旨在实现不仅安全而且有意义的帮助。我们发布了Oy1、代码和基准测试，以支持负责任的、以用户为中心的人工智能。
英文摘要：
Large language models (LLMs) typically deploy safety mechanisms to prevent harmful content generation. Most current approaches focus narrowly on risks posed by malicious actors, often framing risks as adversarial events and relying on defensive refusals. However, in real-world settings, risks also come from non-malicious users seeking help while under psychological distress (e.g., self-harm intentions). In such cases, the model's response can strongly influence the user's next actions. Simple refusals may lead them to repeat, escalate, or move to unsafe platforms, creating worse outcomes. We introduce Constructive Safety Alignment (CSA), a human-centric paradigm that protects against malicious misuse while actively guiding vulnerable users toward safe and helpful results. Implemented in Oyster-I (Oy1), CSA combines game-theoretic anticipation of user reactions, fine-grained risk boundary discovery, and interpretable reasoning control, turning safety into a trust-building process. Oy1 achieves state-of-the-art safety among open models while retaining high general capabilities. On our Constructive Benchmark, it shows strong constructive engagement, close to GPT-5, and unmatched robustness on the Strata-Sword jailbreak dataset, nearing GPT-o1 levels. By shifting from refusal-first to guidance-first safety, CSA redefines the model-user relationship, aiming for systems that are not just safe, but meaningfully helpful. We release Oy1, code, and the benchmark to support responsible, user-centered AI.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
