SOMA：用 System One 模型持续监控推理模型的思考过程
SOMA: Continuous Monitoring and Intervention for System Two Reasoning with System One Models
让小模型当监工盯着大模型思考，超了就提醒或接管，省 36% 推理 token，弃答也更准，不用训练。
论文提出 SOMA（System One Metacognitive Awareness）框架，让 Jev 这类 System One 决策模型在 System Two 推理模型生成过程中持续监控并适时干预。SOMA 分三层：场景层用自然语言描述目标和动作，觉察层由 S1 判断行为是否出现，控制层先提醒 S2，行为持续才接管。跨任务共享一个 S1 模型和一个控制器，无需训练。在过度思考场景下，推理 token 减少最多 36% 且平均准确率仍有提升；在不可回答问题上，弃答率从 70.5% 提高到 87.2%。
SOMA: Continuous Monitoring and Intervention for System Two Reasoning with System One Models
Reasoning models often run into problems without noticing them, even when symptoms are visible in their own text. In humans, metacognition separates monitoring from control: fast, intuitive signals keep watch while slow, deliberate reasoning unfolds and help decide whether to continue, change course, or stop. System One (S1) decision models, such as Jev, are a natural candidate for this role: they return a calibrated probability for a natural-language question about a text in one fast call, yet they are used only as gates before reasoning. We introduce SOMA (System One Metacognitive Awareness), in which an S1 model continuously monitors a System Two (S2) reasoning model and intervenes when needed. SOMA has three layers: a scenario layer states each objective in natural language with predefined actions, an awareness layer lets S1 judge whether the described behavior appears, and a control layer accumulates these judgments, first reminds S2, and takes over only if the behavior persists. Sharing one S1 model and one controller across tasks, without training, SOMA addresses two different failures on two reasoning models: on overthinking, it cuts reasoning tokens by up to 36\% while improving average accuracy, and on unanswerable questions, it raises abstention from 70.5\% to 87.2\%.