Step-On-Feet Tuning: Scaling Self-Alignment of LLMs via Bootstrapping

doi:doi:10.57702/geepzxt9

Step-On-Feet Tuning: Scaling Self-Alignment of LLMs via Bootstrapping

Self-alignment is an effective way to reduce the cost of human annotation while ensuring promising model capability. This objective can be achieved from three aspects: (i) high quality data generation aspect, (ii) ready-made data utilizing aspect, (iii) model internal capacity utilizing aspect.

BibTex:

@dataset{Haoyu_Wang_and_Guozheng_Ma_and_Ziqiao_Meng_and_Zeyu_Qin_and_Li_Shen_and_Zhong_Zhang_and_Bingzhe_Wu_and_Liu_Liu_and_Yatao_Bian_and_Tingyang_Xu_and_Xueqian_Wang_and_Peilin_Zhao_2024,
    abstract = {Self-alignment is an effective way to reduce the cost of human annotation while ensuring promising model capability. This objective can be achieved from three aspects: (i) high quality data generation aspect, (ii) ready-made data utilizing aspect, (iii) model internal capacity utilizing aspect.},
    author = {Haoyu Wang and Guozheng Ma and Ziqiao Meng and Zeyu Qin and Li Shen and Zhong Zhang and Bingzhe Wu and Liu Liu and Yatao Bian and Tingyang Xu and Xueqian Wang and Peilin Zhao},
    doi = {10.57702/geepzxt9},
    institution = {No Organization},
    keyword = {'LLMs', 'bootstrapping', 'data diversity', 'self-alignment'},
    month = {dec},
    publisher = {TIB},
    title = {Step-On-Feet Tuning: Scaling Self-Alignment of LLMs via Bootstrapping},
    url = {https://service.tib.eu/ldmservice/dataset/step-on-feet-tuning--scaling-self-alignment-of-llms-via-bootstrapping},
    year = {2024}
}