This repository is the basic schemas, antonomasias, and the knowledge of relationships referred to in RoCar: A Relationship Network-based Evaluation Method to Large Language Models.
antonomasia.csv
is the antonomasias, in which two columns donate the antonomasia and the gender of a fictional character.name, gender 小红, 0 昊昊, 1
pre_prompts.txt
contains the basic relationships of first-order and second-order.父亲和儿子存在基因传承的父子关系,父亲比儿子大一辈,儿子称呼父亲为爸爸。 父亲的父亲和父亲的儿子是祖孙关系,祖父比孙辈大两辈,孙辈应向祖父称呼爷爷。
derivation_rules.txt
states the common rules for deriving complex relationships.辈分长自己一辈或者两辈则已自己的父母进行类比称呼: 长一辈,例如:妈妈的表哥表弟一表舅、爸爸的表姐妹一表姑姑或表姑妈 长两辈,例如:妈妈的舅舅一舅姥爷、爸爸的姑姑一老姑 长三辈,以自己父母类比加个“太”字例如:爸爸的爷爷一太爷爷、爸爸的外公一太外公、妈妈的外婆一太外婆
task_graph.txt
is the randomly constructed task graph in our experiments.reasoning.txt
contains evaluation tasks in reasoning capability evaluation experiments.memory.txt
contains evaluation tasks in memory capability evaluation experiments.
This was supported in part by the Data Mining Group at Northeastern University. Thanks to OpenAI, Anthropic, Baidu, Xunfei, and Tsinghua University for providing LLM applications and documentation support.
@misc{wang2023rocar,
title={RoCar: A Relationship Network-based Evaluation Method to Large Language Models},
author={Ming Wang and Wenfang Wu and Chongyun Gao and Daling Wang and Shi Feng and Yifei Zhang},
year={2023},
eprint={2307.15997},
archivePrefix={arXiv},
primaryClass={cs.CL}
}
- More types of entities and relationships are in the task graph.
- Multi-group random experiments.
- More colorful prompts.