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 2404.03134cs.CLcs.CY
收藏

英语大语言模型的鲁棒代词保真度:它们是在推理、重复还是仅存在偏见?

Robust Pronoun Fidelity with English LLMs: Are they Reasoning, Repeating, or Just Biased?

Saarland University(萨尔大学);University of Hamburg(汉堡大学)

Vagrant Gautam, Eileen Bingert, Dawei Zhu, Anne Lauscher, Dietrich Klakow

AI总结:

该研究提出代词保真度任务,构建含超500万实例的RUFF数据集,评估37个不同架构与规模的英语大语言模型,发现模型代词保真度鲁棒性不足,易受干扰,呼吁关注相关差距与合理评估模型推理能力。

摘要:

随着语言模型的应用日益广泛,为个体提供鲁棒、忠实且无危害的代词使用是语言模型开发的重要目标,但以往研究往往仅同时探究其中一到两项特性。为衡量在这一综合目标上的进展,我们提出了代词保真度任务:给定一个引入了共指实体和代词的上下文,该任务要求在后续内容中复用正确的代词。我们构建了RUFF,这是一个经过精心设计、包含超500万个实例的数据集,用于衡量英语中的鲁棒代词保真度;我们评估了来自9个主流模型家族的37个模型变体,覆盖不同架构(仅编码器、仅解码器、编码器-解码器)和参数规模(11M-70B)。当个体随代词被引入时,模型大多能在下一句中忠实地复用该代词,但在处理she/her/her、单数they及新代词时表现明显更差。此外,模型很容易被讨论其他人物的非对抗性句子干扰;哪怕仅一句带有干扰代词的句子,都会使准确率平均下降34个百分点。我们的结果表明,在人类能达到近100%准确率的简单自然场景中,代词保真度并不具备鲁棒性。我们鼓励研究人员弥补我们发现的这些差距,并在表层重复可能夸大模型性能感知的场景中谨慎评估推理能力。

英文摘要:

Robust, faithful and harm-free pronoun use for individuals is an important goal for language model development as their use increases, but prior work tends to study only one or two of these characteristics at a time. To measure progress towards the combined goal, we introduce the task of pronoun fidelity: given a context introducing a co-referring entity and pronoun, the task is to reuse the correct pronoun later. We present RUFF, a carefully-designed dataset of over 5 million instances to measure robust pronoun fidelity in English, and we evaluate 37 model variants from nine popular families, across architectures (encoder-only, decoder-only and encoder-decoder) and scales (11M-70B parameters). When an individual is introduced with a pronoun, models can mostly faithfully reuse this pronoun in the next sentence, but they are significantly worse with she/her/her, singular they and neopronouns. Moreover, models are easily distracted by non-adversarial sentences discussing other people; even one sentence with a distractor pronoun causes accuracy to drop on average by 34 percentage points. Our results show that pronoun fidelity is not robust, in a simple, naturalistic setting where humans achieve nearly 100% accuracy. We encourage researchers to bridge the gaps we find and to carefully evaluate reasoning in settings where superficial repetition might inflate perceptions of model performance.

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。

↑