@article{largelanguagemodelsstruggletodescribe,
  title = {Large Language Models Struggle to Describe the Haystack without Human Help: Human-in-the-loop Evaluation of LLMs},
  author = {Zongxia Li and Lorena Calvo-Bartolomé and Alexander Hoyle and Paiheng Xu and Alden Dima and Juan Francisco Fung and Jordan Boyd-Graber},
  year = {2025},
  eprint = {2502.14748},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2502.14748v1},
}