Abstract
The goal of ALTA 2023 Shared Task is to distinguish between human-authored text and synthetic text generated by Large Language Models (LLMs). Given the growing societal concerns surrounding LLMs, this task addresses the urgent need for robust text verification strategies. In this paper, we describe our method, a fine-tuned Falcon-7B model with incorporated label smoothing into the training process. We applied model prompting to samples with lower confidence scores to enhance prediction accuracy. Our model achieved a statistically significant accuracy of 0.991.- Anthology ID:
- 2023.alta-1.18
- Volume:
- Proceedings of the 21st Annual Workshop of the Australasian Language Technology Association
- Month:
- November
- Year:
- 2023
- Address:
- Melbourne, Australia
- Editors:
- Smaranda Muresan, Vivian Chen, Kennington Casey, Vandyke David, Dethlefs Nina, Inoue Koji, Ekstedt Erik, Ultes Stefan
- Venue:
- ALTA
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 153–158
- Language:
- URL:
- https://aclanthology.org/2023.alta-1.18
- DOI:
- Cite (ACL):
- Rinaldo Gagiano and Lin Tian. 2023. A Prompt in the Right Direction: Prompt Based Classification of Machine-Generated Text Detection. In Proceedings of the 21st Annual Workshop of the Australasian Language Technology Association, pages 153–158, Melbourne, Australia. Association for Computational Linguistics.
- Cite (Informal):
- A Prompt in the Right Direction: Prompt Based Classification of Machine-Generated Text Detection (Gagiano & Tian, ALTA 2023)
- PDF:
- https://preview.aclanthology.org/ml4al-ingestion/2023.alta-1.18.pdf