Files

ModelHub XC f9d9d3a8a6 初始化项目，由ModelHub XC社区提供模型

Model: pymlex/Qwen2.5-0.5B-Human
Source: Original Platform

2026-07-07 20:13:26 +08:00

4.0 KiB

Raw Permalink Blame History

license, language, base_model, tags, library_name, datasets, metrics, pipeline_tag

license

language

base_model

Qwen2.5-0.5B-Human

DPO fine-tune of Qwen/Qwen2.5-0.5B-Instruct that paraphrases Spanish academic abstracts to reduce AI-detection scores from danibor/oculus-v2.0-multilingual.

Overview

Preference pairs for optimisation come from pymlex/ai-generated-texts. The corpus is Flaglab/academic-knowledge-abstracts-es. For each train abstract, two base-model paraphrases are ranked by Oculus logit. DPO with beta = 0.1 increases the relative log-probability of the lower-logit completion. Retained pairs: 6396 from 8891 train abstracts with absolute logit gap at least 1.

Evaluation setup

Hardware: NVIDIA RTX 5090, Ubuntu Jupyter, CUDA 13.0+, bf16 training and inference. Post-training evaluation generates one paraphrase per validation and test abstract with the base and fine-tuned models, scores each output with Oculus, and treats label 1 as AI-generated at threshold 0.5 on detector probability.

During DPO, mean validation AI probability on a 276-text subset moved from 0.6740 at step 0 to 0.2437 at the last monitor step (-0.4303).

Results

Model	Split	n	mean prob	mean logit	accuracy	ROC-AUC	F1
base	validation	1107	0.6550	1.4619	0.6712	n/a	0.8032
base	test	1112	0.6532	1.5581	0.6655	n/a	0.7991
fine-tuned	validation	1107	0.2264	-2.0100	0.1716	n/a	0.2930
fine-tuned	test	1112	0.2391	-1.8733	0.1835	n/a	0.3100

Lower mean probability and MCC near zero indicate weaker detector response on model paraphrases under the AI-positive labelling convention.

Source code

The full pipeline is published on GitHub.

Citation

If you found this model useful, please cite it as:

@misc{zyukov2026qwenhuman,
  title         = {{Qwen2.5-0.5B-Human: DPO fine-tune against Oculus detector}},
  author        = {Zyukov, Alex},
  year          = {2026},
  url           = {https://huggingface.co/pymlex/Qwen2.5-0.5B-Human}
}

@misc{zyukov2026aitexttricking,
  title         = {{DPO Fine-Tuning Against Multilingual AI Text Detectors}},
  author        = {Zyukov, Alex},
  year          = {2026},
  url           = {https://github.com/pymlex/ai-text-detector-tricking},
  publisher     = {GitHub},
  organization  = {pymlex}
}

@misc{nicks2024detectors,
  title         = {{Language Model Detectors Are Easily Optimized Against}},
  author        = {Nicks, Cameron and Chua, Jeremy and Liu, Stephen and others},
  year          = {2024},
  eprint        = {2406.07490},
  archivePrefix = {arXiv},
  primaryClass  = {cs.CL},
  url           = {https://arxiv.org/abs/2406.07490}
}

@misc{oculus2026,
  title         = {{Oculus 2.0 Multilingual AI Text Detector}},
  author        = {danibor},
  year          = {2026},
  url           = {https://huggingface.co/danibor/oculus-v2.0-multilingual}
}

@misc{flaglab2025abstracts,
  title         = {{Academic Knowledge Abstracts Spanish}},
  author        = {Flaglab},
  year          = {2025},
  url           = {https://huggingface.co/datasets/Flaglab/academic-knowledge-abstracts-es}
}

The project is under GPL-3.0 license.

4.0 KiB Raw Permalink Blame History