通过语义感知排列训练缓解大型语言模型中的反转诅咒
Mitigating Reversal Curse in Large Language Models via Semantic-aware Permutation Training
AI总结：
针对因果LLMs因词序差异导致的反转诅咒问题，提出语义感知排列训练（SPT），利用辅助模型将句子分割为语义单元再进行排列，使模型有效预测前文，显著提升了双向推理性能。
摘要：
尽管大型语言模型（LLMs）在各种任务中取得了令人印象深刻的性能，但最近的研究表明因果LLMs存在“反转诅咒”。一个典型的例子是，模型知道“A的父亲是B”，却无法推理出“B的孩子是A”。这一局限性对通用人工智能（AGI）的发展构成了挑战，因为它表明模型在理解和应用双向推理的能力上存在差距。在本文中，我们首先进行了大量评估，并发现反转诅咒的根本原因在于训练和推理阶段之间的词序不同，即因果语言模型预测训练数据中前文词汇的能力较差。因此，对训练数据进行排列被视为一种潜在的解决方案，因为这能让模型预测前文词汇或标记。然而，以往的排列方法可能会破坏完整的短语或实体，从而给模型理解和学习训练数据带来挑战。为了解决这个问题，我们提出了语义感知排列训练（SPT），该方法通过使用辅助语言模型将训练句子分割成语义单元（即实体或短语），并在输入模型前对这些单元进行排列来解决此问题。大量实验表明，SPT有效缓解了反转诅咒，因为在反向问题上的性能接近于正向问题，并显著提升了现有工作的性能。
英文摘要：
While large language models (LLMs) have achieved impressive performance across diverse tasks, recent studies showcase that causal LLMs suffer from the "reversal curse". It is a typical example that the model knows "A's father is B", but is unable to reason "B's child is A". This limitation poses a challenge to the advancement of artificial general intelligence (AGI), as it suggests a gap in the models' ability to comprehend and apply bidirectional reasoning. In this paper, we first conduct substantial evaluation and identify that the root cause of the reversal curse lies in the different word order between the training and inference stage, namely, the poor ability of causal language models to predict antecedent words within the training data. Accordingly, permutation on the training data is considered as a potential solution, since this can make the model predict antecedent words or tokens. However, previous permutation methods may disrupt complete phrases or entities, thereby posing challenges for the model to comprehend and learn from training data. To address this issue, we propose Semantic-aware Permutation Training (SPT), which addresses this issue by segmenting the training sentences into semantic units (i.e., entities or phrases) with an assistant language model and permuting these units before feeding into the model. Extensive experiments demonstrate that SPT effectively mitigates the reversal curse since the performance on reversed questions approximates that on the forward ones, and significantly advances the performance of existing works.
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。
