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 2409.13843cs.CLcs.AIcs.CY
收藏

STOP!基于冒犯性递进的敏感性测试对大型语言模型进行基准测试

STOP! Benchmarking Large Language Models with Sensitivity Testing on Offensive Progressions

Brock University(布鲁克大学)

Robert Morabito, Sangmitra Madhusudan, Tyler McDonald, Ali Emami

AI总结:

本文提出STOP数据集,通过450个逐步升级的冒犯性递进序列评估LLMs的偏见检测能力,发现现有模型表现不一致,且基于人类判断对齐模型能显著提升其在敏感任务上的表现。

摘要:

缓解大型语言模型(LLMs)中的显式和隐式偏见已成为自然语言处理领域的关键重点。然而,当前的许多方法孤立地评估场景,未考虑更广泛的背景或每种情况中潜在偏见的范围。为此,我们引入了冒犯性递进敏感性测试(STOP)数据集,其中包含450个冒犯性递进序列,共计2700个严重程度不一的独立句子,这些句子从较轻逐步升级为更明显的冒犯。STOP覆盖了9个人口统计群体和46个子群体的广泛范围,确保了包容性和全面覆盖。我们评估了包括GPT-4、Mixtral和Llama 3在内的多个领先闭源与开源模型。研究结果显示,即使是表现最好的模型在检测偏见时也不一致,成功率从19.3%到69.8%不等。我们还展示了将模型在STOP上与人类判断对齐,如何能将模型在BBQ、StereoSet和CrowS-Pairs等敏感任务上的回答率提高多达191%,同时保持甚至提升性能。STOP为评估LLMs中复杂的偏见本质提出了一个新框架,这将有助于制定更有效的偏见缓解策略并促进更公平语言模型的创建。

英文摘要:

Mitigating explicit and implicit biases in Large Language Models (LLMs) has become a critical focus in the field of natural language processing. However, many current methodologies evaluate scenarios in isolation, without considering the broader context or the spectrum of potential biases within each situation. To address this, we introduce the Sensitivity Testing on Offensive Progressions (STOP) dataset, which includes 450 offensive progressions containing 2,700 unique sentences of varying severity that progressively escalate from less to more explicitly offensive. Covering a broad spectrum of 9 demographics and 46 sub-demographics, STOP ensures inclusivity and comprehensive coverage. We evaluate several leading closed- and open-source models, including GPT-4, Mixtral, and Llama 3. Our findings reveal that even the best-performing models detect bias inconsistently, with success rates ranging from 19.3% to 69.8%. We also demonstrate how aligning models with human judgments on STOP can improve model answer rates on sensitive tasks such as BBQ, StereoSet, and CrowS-Pairs by up to 191%, while maintaining or even improving performance. STOP presents a novel framework for assessing the complex nature of biases in LLMs, which will enable more effective bias mitigation strategies and facilitates the creation of fairer language models.

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。

↑