作者位次 #3/7ACL 2026Poster· Natural Language Processing
Revisiting the Reliability of Language Models in Instruction-Following
Jianshuo Dong · Yutong Zhang · Liu Yan · Zhenyu Zhong · Tao Wei · Chao Zhang · Han Qiu
摘要
Advanced LLMs have achieved near-ceiling instruction-following accuracy on benchmarks such as IFEval. However, these impressive scores do not necessarily translate to reliable services in real-world use, where users often vary their phrasing, contextual framing, and task formulations. In this pap...