DAR: 基于智能体框架的道义推理
DAR: Deontic Reasoning with Agentic Harnesses
AI总结：
提出DAR框架，通过让模型按需与法规交互来提升基于LLM的道义推理能力，实验表明智能体框架可提升性能但存在非均匀改进和弱模型数值任务退化问题。
摘要：
道义推理是通过将明确的规则和政策应用于具体案例事实来回答问题，例如根据法规计算纳税义务或确定移民上诉结果。基于LLM的道义推理的一个关键技术挑战是相关规则集可能很长且相互引用，因此模型可能仍无法找到特定推理步骤所需的规则。我们引入了道义智能体推理（DAR），这是一种智能体推理设置，其中模型按需与法规交互。我们在DeonticBench的困难子集上使用多种框架评估DAR。在这些设置中，我们发现智能体框架可以推动道义推理任务的前沿，但改进并不均匀：较弱的模型在数值任务上往往性能下降，同时消耗更多的令牌。
英文摘要：
Deontic reasoning is the task of answering questions by applying explicit rules and policies to case-specific facts, for example computing tax liability under a statute or determining the outcome of an immigration appeal. A key technical challenge for LLM-based deontic reasoning is that the relevant ruleset can be long and cross-referenced, so models may still fail to locate the rules needed for a particular reasoning step. We introduce Deontic Agentic Reasoning (DAR), an agentic reasoning setup in which the model interacts with the statutes on demand. We evaluate DAR under multiple harnesses on hard subsets of DeonticBench. Across these settings, we find that agentic harnesses can push the frontier on deontic reasoning tasks, but improvements are not uniform: weaker models often degrade on numerical tasks while consuming far more tokens.
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。
