LuxVeri at GenAI Detection Task 3: Cross-Domain Detection of AI-Generated Text Using Inverse Perplexity-Weighted Ensemble of Fine-Tuned Transformer Models

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Mobin, Md Kamrujjaman, Islam, Md Saiful
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866915113031696384
author Mobin, Md Kamrujjaman
Islam, Md Saiful
author_facet Mobin, Md Kamrujjaman
Islam, Md Saiful
contents This paper presents our approach for Task 3 of the GenAI content detection workshop at COLING-2025, focusing on Cross-Domain Machine-Generated Text (MGT) Detection. We propose an ensemble of fine-tuned transformer models, enhanced by inverse perplexity weighting, to improve classification accuracy across diverse text domains. For Subtask A (Non-Adversarial MGT Detection), we combined a fine-tuned RoBERTa-base model with an OpenAI detector-integrated RoBERTa-base model, achieving an aggregate TPR score of 0.826, ranking 10th out of 23 detectors. In Subtask B (Adversarial MGT Detection), our fine-tuned RoBERTa-base model achieved a TPR score of 0.801, securing 8th out of 22 detectors. Our results demonstrate the effectiveness of inverse perplexity-based weighting for enhancing generalization and performance in both non-adversarial and adversarial MGT detection, highlighting the potential for transformer models in cross-domain AI-generated content detection.
format Preprint
id arxiv_https___arxiv_org_abs_2501_11918
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LuxVeri at GenAI Detection Task 3: Cross-Domain Detection of AI-Generated Text Using Inverse Perplexity-Weighted Ensemble of Fine-Tuned Transformer Models
Mobin, Md Kamrujjaman
Islam, Md Saiful
Computation and Language
Artificial Intelligence
This paper presents our approach for Task 3 of the GenAI content detection workshop at COLING-2025, focusing on Cross-Domain Machine-Generated Text (MGT) Detection. We propose an ensemble of fine-tuned transformer models, enhanced by inverse perplexity weighting, to improve classification accuracy across diverse text domains. For Subtask A (Non-Adversarial MGT Detection), we combined a fine-tuned RoBERTa-base model with an OpenAI detector-integrated RoBERTa-base model, achieving an aggregate TPR score of 0.826, ranking 10th out of 23 detectors. In Subtask B (Adversarial MGT Detection), our fine-tuned RoBERTa-base model achieved a TPR score of 0.801, securing 8th out of 22 detectors. Our results demonstrate the effectiveness of inverse perplexity-based weighting for enhancing generalization and performance in both non-adversarial and adversarial MGT detection, highlighting the potential for transformer models in cross-domain AI-generated content detection.
title LuxVeri at GenAI Detection Task 3: Cross-Domain Detection of AI-Generated Text Using Inverse Perplexity-Weighted Ensemble of Fine-Tuned Transformer Models
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2501.11918