ConU: Conformal Uncertainty in Large Language Models with Correctness Coverage Guarantees
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866917839780184064 |
|---|---|
| author | Wang, Zhiyuan Duan, Jinhao Cheng, Lu Zhang, Yue Wang, Qingni Shi, Xiaoshuang Xu, Kaidi Shen, Hengtao Zhu, Xiaofeng |
| author_facet | Wang, Zhiyuan Duan, Jinhao Cheng, Lu Zhang, Yue Wang, Qingni Shi, Xiaoshuang Xu, Kaidi Shen, Hengtao Zhu, Xiaofeng |
| contents | Uncertainty quantification (UQ) in natural language generation (NLG) tasks remains an open challenge, exacerbated by the closed-source nature of the latest large language models (LLMs). This study investigates applying conformal prediction (CP), which can transform any heuristic uncertainty notion into rigorous prediction sets, to black-box LLMs in open-ended NLG tasks. We introduce a novel uncertainty measure based on self-consistency theory, and then develop a conformal uncertainty criterion by integrating the uncertainty condition aligned with correctness into the CP algorithm. Empirical evaluations indicate that our uncertainty measure outperforms prior state-of-the-art methods. Furthermore, we achieve strict control over the correctness coverage rate utilizing 7 popular LLMs on 4 free-form NLG datasets, spanning general-purpose and medical scenarios. Additionally, the calibrated prediction sets with small size further highlights the efficiency of our method in providing trustworthy guarantees for practical open-ended NLG applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_00499 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | ConU: Conformal Uncertainty in Large Language Models with Correctness Coverage Guarantees Wang, Zhiyuan Duan, Jinhao Cheng, Lu Zhang, Yue Wang, Qingni Shi, Xiaoshuang Xu, Kaidi Shen, Hengtao Zhu, Xiaofeng Computation and Language Artificial Intelligence Machine Learning Uncertainty quantification (UQ) in natural language generation (NLG) tasks remains an open challenge, exacerbated by the closed-source nature of the latest large language models (LLMs). This study investigates applying conformal prediction (CP), which can transform any heuristic uncertainty notion into rigorous prediction sets, to black-box LLMs in open-ended NLG tasks. We introduce a novel uncertainty measure based on self-consistency theory, and then develop a conformal uncertainty criterion by integrating the uncertainty condition aligned with correctness into the CP algorithm. Empirical evaluations indicate that our uncertainty measure outperforms prior state-of-the-art methods. Furthermore, we achieve strict control over the correctness coverage rate utilizing 7 popular LLMs on 4 free-form NLG datasets, spanning general-purpose and medical scenarios. Additionally, the calibrated prediction sets with small size further highlights the efficiency of our method in providing trustworthy guarantees for practical open-ended NLG applications. |
| title | ConU: Conformal Uncertainty in Large Language Models with Correctness Coverage Guarantees |
| topic | Computation and Language Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2407.00499 |