arXiv：cs.CL· Trieu Hai Nguyen, Sivaswamy Akilesh·· 6 小时前AI 评分34
VietBinoculars：一种零样本检测越南语 LLM 生成文本的方法
VietBinoculars: A Zero-Shot Approach for Detecting Vietnamese LLM-Generated Text
AI 导读
研究提出零样本检测框架 VietBinoculars，将 PhoGPT-4B 的 observer 与 performer 模型与校准后的全局决策阈值结合，并采用越南语专用 BPE 分词，在多领域基准上 AUC 超过 0.99。
正文
Abstract:The rapid proliferation of Large Language Models has intensified the challenge of distinguishing LLM-generated text from human writing in non-English languages. This study introduces VietBinoculars, a zero-shot detection framework coupling PhoGPT-4B observer and performer models with calibrated global decision thresholds. By utilizing specialized Vietnamese BPE tokenization, the method eliminates byte-level fragmentation and probability dilution common in massive multilingual backbones. Evaluated across multi-domain benchmarks, VietBinoculars achieves an area under the ROC curve exceeding 0.99. Under optimal Youden's J thresholds and greedy decoding, detection accuracy reaches at least 98.78\%, while significantly outperforming baseline Binoculars, zero-shot detectors, and commercial tools on creative Capybara prompts. Even under a strict false positive rate constraint of 0.06\%, the detector maintains F1-scores between 83.15\% and 94.70\%. Detection performance consistently improves with sequence length, stabilizing at optimal accuracy for passages containing 450 to 550 tokens. Extended stress testing across 48 distinct model-decoding configurations and three post-generation rewriting strategies delineates practical operational boundaries. VietBinoculars exhibits robust resilience against single-pass paraphrasing and human-style revisions, but experiences notable performance degradation under high-entropy sampling and iterative double paraphrasing.
| Comments: | 39 pages |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2509.26189 [cs.CL] |
| (or arXiv:2509.26189v2 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2509.26189 arXiv-issued DOI via DataCite |
Submission history
From: Trieu Hai Nguyen [view email]
[v1]
Tue, 30 Sep 2025 12:43:37 UTC (1,073 KB)
[v2]
Tue, 6 Oct 2026 15:01:58 UTC (1,759 KB)
来源：arXiv：cs.CL · arxiv.org