CLAMBER: A Benchmark of Identifying and Clarifying Ambiguous Information Needs in Large Language Models
Co-first author: Tong Zhang (Sichuan University), Peixin Qin (Sichuan University)
This is the benchmark test dataset, called CLAMBER, which is used to evaluate LLMs using a well-organized taxonomy in terms of identifying and clarifying ambiguous information needs.
Name | Meaning | Values |
---|---|---|
question | user query | string |
context | context of user query | string |
clarifying_question | suggested clarifying question | string |
require_clarification | If user query is ambiguous | 0/1 |
category | ambiguity type | {"FD": "Epistemic Misalignment", "MC": "Aleatoric Output", "LA": "Linguistic Ambiguity"} |
subclass | sub-type | {"whom": "WHOM", "what": "WHAT", "when": "WHEN", "where": "WHERE", "NK": "UNFAMILIAR", "ICL": "CONTRADICTION", "co-reference": "SEMANTIC", "polysemy": "LEXICAL"} |
If you make advantage of the DREditor in your research, please cite the following in your manuscript:
@misc{zhang2024clamber,
title={CLAMBER: A Benchmark of Identifying and Clarifying Ambiguous Information Needs in Large Language Models},
author={Tong Zhang and Peixin Qin and Yang Deng and Chen Huang and Wenqiang Lei and Junhong Liu and Dingnan Jin and Hongru Liang and Tat-Seng Chua},
year={2024},
booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (ACL)},
}