1. Hyper-multi-step: The Truth Behind Difficult Long-context Tasks
- Author
-
Yu, Yijiong, Xiufa, Ma, Jianwei, Fang, Xu, Zhi, Guangyao, Su, Jiancheng, Wang, Huang, Yongfeng, Qi, Zhixiao, Wang, Wei, Liu, Weifeng, Chen, Ran, and Pei, Ji
- Subjects
Computer Science - Computation and Language - Abstract
Long-context language models (LCLM), characterized by their extensive context window, is becoming increasingly popular. Meanwhile, many long-context benchmarks present challenging tasks that even the most advanced LCLMs struggle to complete. However, the underlying sources of various challenging long-context tasks have seldom been studied. To bridge this gap, we conduct experiments to indicate their difficulty stems primarily from two basic issues: "multi-matching retrieval," which requires the simultaneous retrieval of multiple items, and "logic-based retrieval," which necessitates logical judgment within retrieval criteria. These two problems, while seemingly straightforward, actually exceed the capabilities of LCLMs because they are proven to be hyper-multi-step (demanding numerous steps to solve) in nature. This finding could explain why LLMs struggle with more advanced long-context tasks, providing a more accurate perspective for rethinking solutions for them., Comment: Our code is publicly available at https://github.com/yuyijiong/hard_retrieval_for_llm and the datasets is at https://huggingface.co/datasets/yuyijiong/difficult_retrieval
- Published
- 2024