arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2406.02120cs.CL
收藏

Diver:基于跨度级互信息验证的大型语言模型解码

Diver: Large Language Model Decoding with Span-Level Mutual Information Verification

Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所);School of Artificial Intelligence, University of Chinese Academy of Sciences(中国科学院大学人工智能学院);Wuhan AI Research(武汉人工智能研究院)

Jinliang Lu, Chen Wang, Jiajun Zhang

AI总结:

本文提出Diver方法,通过跨度级逐点互信息(PMI)验证来增强大型语言模型解码,在推理时识别发散步骤并基于对数似然增益计算PMI分数以选择最佳候选跨度,有效解决输出偏离输入的问题并在性能和多功能性上显著优于现有方法。

摘要:

大型语言模型(LLMs)在提供特定任务指令时,展现出了适应各种任务的出色能力。然而,使用标准解码策略的LLMs经常会在偏离输入方面遇到困难。直观地说,合规的LLM输出应该反映输入中存在的信息,这可以通过逐点互信息(PMI)分数来衡量。因此,我们提出了Diver,一种通过跨度级PMI验证来增强LLM解码的新方法。在推理过程中,Diver首先识别可能导致多个候选跨度的发散步骤。随后,它通过评估如果生成候选跨度时输入的对数似然增益来计算PMI分数。最后,基于PMI重新排序的输出分布选择最佳跨度。我们在各种下游任务上评估了我们的方法,实证结果表明,Diver在性能和多功能性上均显著优于现有的解码方法。

英文摘要:

Large language models (LLMs) have shown impressive capabilities in adapting to various tasks when provided with task-specific instructions. However, LLMs using standard decoding strategies often struggle with deviations from the inputs. Intuitively, compliant LLM outputs should reflect the information present in the input, which can be measured by point-wise mutual information (PMI) scores. Therefore, we propose Diver, a novel approach that enhances LLM Decoding through span-level PMI verification. During inference, Diver first identifies divergence steps that may lead to multiple candidate spans. Subsequently, it calculates the PMI scores by assessing the log-likelihood gains of the input if the candidate spans are generated. Finally, the optimal span is selected based on the PMI re-ranked output distributions. We evaluate our method across various downstream tasks, and empirical results demonstrate that Diver significantly outperforms existing decoding methods in both performance and versatility.

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。

↑