用于忠实且抽象的对话生成的弹性权重移除
Elastic Weight Removal for Faithful and Abstractive Dialogue Generation
AI总结：
提出弹性权重移除(EWR)方法，利用Fisher信息矩阵权衡参数重要性以抑制对话幻觉，并扩展为同时抑制幻觉与抽取性回复，提升生成忠实性。
摘要：
理想情况下，对话系统生成的回复应忠实于相关文档中包含的知识。然而，许多模型会生成幻觉回复，与文档相矛盾或包含无法验证的信息。为了缓解这种不良行为，已有方法提出在负样本上微调一个“负专家”，并将其参数从预训练模型的参数中减去。然而，直觉上，这没有考虑到某些参数在导致幻觉方面比其他参数负有更大的责任。因此，我们提出通过 Fisher Information 矩阵（的近似）来权衡它们各自的重要性，该矩阵衡量了其估计的不确定性。我们将此方法称为 Elastic Weight Removal (EWR)。我们使用不同变体的 Flan-T5 作为骨干语言模型，在多个信息寻求对话生成数据集上评估了我们的方法，并将该方法与 CTRL、Quark、DExperts 和 Noisy Channel reranking 等最先进的忠实性技术进行了比较。广泛的自动和人工评估表明，EWR 系统性地提高了忠实性，而在其他指标上仅有微小的代价。然而，我们注意到，仅仅抑制幻觉可能会增加抽取性，即对文档片段的浅层复制粘贴，这可能是不理想的。因此，作为第二个主要贡献，我们展示了我们的方法可以扩展为同时抑制幻觉和抽取性回复。我们公开发布了用于复现 EWR 和所有基线的代码。
英文摘要：
Ideally, dialogue systems should generate responses that are faithful to the knowledge contained in relevant documents. However, many models generate hallucinated responses instead that contradict it or contain unverifiable information. To mitigate such undesirable behaviour, it has been proposed to fine-tune a `negative expert' on negative examples and subtract its parameters from those of a pre-trained model. However, intuitively, this does not take into account that some parameters are more responsible than others in causing hallucinations. Thus, we propose to weigh their individual importance via (an approximation of) the Fisher Information matrix, which measures the uncertainty of their estimate. We call this method Elastic Weight Removal (EWR). We evaluate our method -- using different variants of Flan-T5 as a backbone language model -- on multiple datasets for information-seeking dialogue generation and compare our method with state-of-the-art techniques for faithfulness, such as CTRL, Quark, DExperts, and Noisy Channel reranking. Extensive automatic and human evaluation shows that EWR systematically increases faithfulness at minor costs in terms of other metrics. However, we notice that only discouraging hallucinations may increase extractiveness, i.e. shallow copy-pasting of document spans, which can be undesirable. Hence, as a second main contribution, we show that our method can be extended to simultaneously discourage hallucinations and extractive responses. We publicly release the code for reproducing EWR and all baselines.
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。
