可解码但不忠实：将自然语言理由与程序化验证器耦合
Decodable but Not Faithful: Coupling Natural-Language Rationales to Programmatic Verifiers
AI总结：
提出验证器耦合推理框架，通过一致性训练使验证器信息可从理由表示中解码，但发现可解码性不保证忠实生成，在多个任务中验证了该差距。
摘要：
语言模型可以为其预测生成看似合理的理由，但这些解释可能并不忠实地代表模型的内部推理。我们提出了验证器耦合推理框架，该框架在推理轨迹中插入内联声明，并训练一个辅助一致性头来从理由跨度隐藏状态预测程序化验证器输出。核心发现是可解码性与忠实性之间存在差距：一致性训练可靠地使验证器信息可从理由表示中解码，但可解码性并不保证忠实生成。在LeanCheck（形式定理证明）中，仅理由和仅证明池化在反事实冲突下实现了完美的方向分离。在KataGo（围棋引擎）中，评论跨度以81%的准确率编码10路胜率区间。然而，在代码设置中，模型实现了98.6%的耦合，但其生成的解释仍然不忠实：流畅的散文带有正确的结构化声明，但描述了无关的算法；一个受控的预训练与从头开始比较表明，这种差距并非由能力驱动。合成激活修补证实了因果影响（73-89%对比31%基线），FEVER显示仅证据池化以原始准确性为代价隔离了真正的证据敏感性，而逐声明分析表明一致性损失对细粒度声明的益处大于二元声明。这些结果表明，一致性损失是有效的诊断和表示塑造工具，但不是忠实推理的充分条件。
英文摘要：
Language models can generate plausible rationales for their predictions, but these explanations may not faithfully represent the model's internal reasoning. We propose verifier-coupled reasoning, a framework that inserts inline claims into reasoning traces and trains an auxiliary consistency head to predict programmatic verifier outputs from rationale-span hidden states. The central finding is a gap between decodability and faithfulness: consistency training reliably makes verifier information decodable from rationale representations, but decodability does not guarantee faithful generation. In LeanCheck (formal theorem proving), rationale-only and proof-only pooling achieve perfect directional separation under counterfactual conflict. In KataGo (Go engine), commentary spans encode 10-way win-rate buckets at 81% accuracy. Yet in a code setting, the model achieves 98.6% coupling while its generated explanations remain unfaithful: fluent prose with correct structured claims, but describing unrelated algorithms; a controlled pretrained-vs-from-scratch comparison shows the gap is not capacity-driven. Synthetic activation patching confirms causal influence (73-89% vs. 31% baseline), FEVER reveals that evidence-only pooling isolates genuine evidence sensitivity at the cost of raw accuracy, and per-claim analysis shows that consistency loss disproportionately benefits fine-grained claims over binary ones. These results establish that consistency losses are effective diagnostics and representation-shaping tools, but not sufficient conditions for faithful reasoning.
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。
