何时相信廉价检查：弱验证与强验证用于推理
When to Trust the Cheap Check: Weak and Strong Verification for Reasoning
AI总结：
本文提出弱-强验证策略，通过两个阈值结构控制接受和拒绝错误，以提高LLM推理的可靠性和准确性。
摘要：
在大型语言模型（LLM）推理过程中，越来越多地融入了更广泛的验证循环。内部，系统使用廉价检查，如自一致性或代理奖励，我们称之为弱验证。外部，用户检查输出并通过反馈引导模型，直到结果可信，这称为强验证。这些信号在成本和可靠性上存在显著差异：强验证可以建立信任但资源消耗大，而弱验证速度快且可扩展但噪声大且不完美。我们通过弱-强验证策略正式化这种张力，这些策略决定何时接受或拒绝基于弱验证，何时推迟到强验证。我们引入了捕捉错误接受、错误拒绝和强验证频率的指标。在总体上，我们证明最优策略具有两个阈值结构，并且校准和锐度决定了弱验证器的价值。基于此，我们开发了一个在线算法，可以保证控制接受和拒绝错误，而无需对查询流、语言模型或弱验证器做出假设。
英文摘要：
Reasoning with LLMs increasingly unfolds inside a broader verification loop. Internally, systems use cheap checks, such as self-consistency or proxy rewards, which we call weak verification. Externally, users inspect outputs and steer the model through feedback until results are trustworthy, which we call strong verification. These signals differ sharply in cost and reliability: strong verification can establish trust but is resource-intensive, while weak verification is fast and scalable but noisy and imperfect. We formalize this tension through weak--strong verification policies, which decide when to accept or reject based on weak verification and when to defer to strong verification. We introduce metrics capturing incorrect acceptance, incorrect rejection, and strong-verification frequency. Over population, we show that optimal policies admit a two-threshold structure and that calibration and sharpness govern the value of weak verifiers. Building on this, we develop an online algorithm that provably controls acceptance and rejection errors without assumptions on the query stream, the language model, or the weak verifier.
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。
