Prompting from the bench: Large-scale pretraining is not sufficient to prepare LLMs for ordinary meaning analysis
从bench出发的提示:大规模预训练不足以使LLM为普通意义分析做准备
机构 * Georgetown University(乔治城大学) ; University of South Carolina(南卡罗来纳大学)
专题命中 预训练与数据 :LLM(title_cn,summary_cn);pretraining(title);prompting(title);large language model(abstract)
AI总结 本文通过实验证明,LLM在法律文本的普通意义分析中存在鲁棒性不足的问题,质疑其在实际应用中的有效性。
Comments Accepted FAccT 2026; 29 pages, 14 tables, 7 figures. Previous title - Not ready for the bench: LLM legal interpretation is unstable and out of step with human judgments; NLLPW 2026