arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2610.39016cs.CLcs.AI
收藏

Transformer中的隐式推理:弥合类GPT-2模型的组合性差距

Implicit Reasoning in Transformers: Bridging the Compositionality Gap in GPT-2-like Models

Davide Venuto, Leonardo Venuto

AI总结:

该研究针对类GPT-2模型的组合性差距,通过调整训练分布的原子事实呈现格式,成功弥合了该差距,证明其可通过机制导向的训练设计克服。

摘要:

组合泛化仍然是Transformer语言模型的一个持续存在的局限性:即使模型在分布内(ID)正确学习了组合已知事实,当将相同的组合机制应用于分布外(OOD)知识时,也可能失败。我们使用从头训练的类GPT-2 Transformer模型,在受控的两跳推理场景中研究这种组合性差距。基于先前识别出与组合泛化相关的回路的工作,我们测试OOD迁移的失败是否与训练期间原子知识的结构呈现方式有关。我们修改训练分布,以与学习到的组合回路的第二跳检索步骤位置对齐的格式呈现ID和OOD原子事实,而不修改模型架构或添加OOD组合监督。在这种干预下,模型成功将学习到的ID组合行为迁移到保留的OOD原子事实,在受控场景中弥合了观察到的组合性差距。我们的结果提供了证据,表明观察到的组合性差距不是基本的架构局限性,并且可以通过基于机制的训练分布设计来克服。

英文摘要:

Compositional generalization remains a persistent limitation of transformer language models: even when a model learns to combine known facts correctly in-distribution (ID), the same compositional mechanism may fail when applied to out-of-distribution (OOD) knowledge. We investigate this compositionality gap in a controlled two-hop reasoning setting using a GPT-2-like transformer model trained from scratch. Building on prior work identifying a circuit associated with compositional generalization, we test whether the failure of OOD transfer is related to how atomic knowledge is structurally exposed during training. We modify the training distribution by presenting both ID and OOD atomic facts in a format positionally aligned with the second-hop retrieval step of the learned compositional circuit, without modifying the model architecture or adding OOD compositional supervision. Under this intervention, the model successfully transfers the learned ID compositional behaviour to held-out OOD atomic facts, closing the observed compositionality gap in the controlled setting. Our results provide evidence that the observed compositionality gap is not a fundamental architecture limitation and that it can be overcome through mechanistically informed training distribution design.

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。

↑