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 2401.06853cs.CL
收藏

大语言模型能够学习时间推理

Large Language Models Can Learn Temporal Reasoning

Georgia Institute of Technology(佐治亚理工学院);Cisco Research(思科研究院)

Siheng Xiong, Ali Payani, Ramana Kompella, Faramarz Fekri

AI总结:

针对大语言模型在时间推理上的难题,提出TG-LLM框架,借助时间图表示、TGQA合成数据集微调及CoT自举等策略,实现可迁移的时间推理能力并提升推理可靠性。

摘要:

尽管大语言模型(LLM)已展现出卓越的推理能力,但它们仍存在缺陷与不准确之处。近期研究提出了多种方法来缓解这些局限。其中时间推理(TR)因依赖多样的时间概念与复杂的时间逻辑,对LLM构成了重大挑战。本文提出了TG-LLM,一种面向基于语言的时间推理的新型框架。我们未在原始上下文上进行推理,而是采用了一种潜在表示——时间图(TG),以增强时间推理的学习效果。我们构建了一个完全可控、所需监督极少的合成数据集TGQA,用于针对文本到时间图转换任务微调LLM。实验证实,在该数据集上学到的时间图转换能力可迁移至其他时间推理任务与基准。除此之外,我们通过思维链(CoT)自举法与图数据增强,教LLM在时间图上进行审慎推理。我们观察到,这些在实用性与多样性间保持平衡的策略,相比 vanilla CoT 蒸馏能带来更可靠的思维链与最终结果。

英文摘要:

While large language models (LLMs) have demonstrated remarkable reasoning capabilities, they are not without their flaws and inaccuracies. Recent studies have introduced various methods to mitigate these limitations. Temporal reasoning (TR), in particular, presents a significant challenge for LLMs due to its reliance on diverse temporal concepts and intricate temporal logic. In this paper, we propose TG-LLM, a novel framework towards language-based TR. Instead of reasoning over the original context, we adopt a latent representation, temporal graph (TG) that enhances the learning of TR. A synthetic dataset (TGQA), which is fully controllable and requires minimal supervision, is constructed for fine-tuning LLMs on this text-to-TG translation task. We confirmed in experiments that the capability of TG translation learned on our dataset can be transferred to other TR tasks and benchmarks. On top of that, we teach LLM to perform deliberate reasoning over the TGs via Chain-of-Thought (CoT) bootstrapping and graph data augmentation. We observed that those strategies, which maintain a balance between usefulness and diversity, bring more reliable CoTs and final results than the vanilla CoT distillation.

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。

↑