arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2305.14540cs.CL
收藏

LLMs作为事实推理者:来自现有基准的洞察与超越

LLMs as Factual Reasoners: Insights from Existing Benchmarks and Beyond

Salesforce AI(Salesforce 人工智能研究院)

Philippe Laban, Wojciech Kryściński, Divyansh Agarwal, Alexander R. Fabbri, Caiming Xiong, Shafiq Joty, Chien-Sheng Wu

AI总结:

本文提出创建事实不一致检测基准的新协议及10领域基准SummEdits,发现多数LLMs在复杂任务上表现接近随机,最佳模型GPT-4仍比人类低8%,揭示了LLMs事实推理能力的差距。

摘要:

随着LLMs近期在实际场景中的出现,拥有能有效检测事实不一致的方法对于减少错误信息的传播和提高对模型输出的信任至关重要。在现有的事实一致性基准上进行测试时,我们发现,与传统的非LLM方法相比,少数大型语言模型(LLMs)在事实不一致检测的分类基准上表现具有竞争力。然而,更深入的分析表明,大多数LLMs在该任务更复杂的表述上会失败,并暴露出现有评估基准的问题,从而影响评估精度。为了解决这个问题,我们提出了一种用于不一致检测基准创建的新协议,并在一个名为SummEdits的10领域基准中实现了它。这个新基准每样本的成本效益比以前的基准高20倍,并且高度可复现,因为我们估计标注者间的一致性约为0.9。大多数LLMs在SummEdits上表现挣扎,性能接近随机概率。表现最好的模型GPT-4仍然比估计的人类表现低8%,这突显了LLMs在事实推理和检测发生时不一致方面的差距。

英文摘要:

With the recent appearance of LLMs in practical settings, having methods that can effectively detect factual inconsistencies is crucial to reduce the propagation of misinformation and improve trust in model outputs. When testing on existing factual consistency benchmarks, we find that a few large language models (LLMs) perform competitively on classification benchmarks for factual inconsistency detection compared to traditional non-LLM methods. However, a closer analysis reveals that most LLMs fail on more complex formulations of the task and exposes issues with existing evaluation benchmarks, affecting evaluation precision. To address this, we propose a new protocol for inconsistency detection benchmark creation and implement it in a 10-domain benchmark called SummEdits. This new benchmark is 20 times more cost-effective per sample than previous benchmarks and highly reproducible, as we estimate inter-annotator agreement at about 0.9. Most LLMs struggle on SummEdits, with performance close to random chance. The best-performing model, GPT-4, is still 8\% below estimated human performance, highlighting the gaps in LLMs' ability to reason about facts and detect inconsistencies when they occur.

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。

↑