Metraipolis › archive › dc_ca804652cb9a

Linear Probing Provides Robust and Efficient Detection of Machine-Generated Text — arXiv:2608.24780

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.24780
License (as archived): CC0-1.0
Archived: 2026-10-18T06:18:18.974Z
SHA-256 of the archived text: 51f81858c4d1d0376a32a718411bca61ea2ba0f047f0f3592bdce45a3b6c73c2
Stable URL of this page: https://metraipolis.com/documento/dc_ca804652cb9a

How to cite

Linear Probing Provides Robust and Efficient Detection of Machine-Generated Text — arXiv:2608.24780 — arXiv (en), https://arxiv.org/abs/2608.24780 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ca804652cb9a (Metraipolis archive dc_ca804652cb9a, sha256 51f81858c4d1d0376a32a718411bca61ea2ba0f047f0f3592bdce45a3b6c73c2)

BibTeX

@misc{dcca804652cb9a,
  title = {Linear Probing Provides Robust and Efficient Detection of Machine-Generated Text — arXiv:2608.24780},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.24780},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ca804652cb9a; sha256 51f81858c4d1d0376a32a718411bca61ea2ba0f047f0f3592bdce45a3b6c73c2},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.24780. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Distinguishing machine-generated text (MGT) from human-written text (HWT) becomes increasingly important due to potential misuse. However, most supervised detectors often degrade out-of-domain (OOD) and require large, diverse training sets. In this work, we analyze the linearity and quality of MGT representations and show that simple linear probes outperform a wide range of detectors while being substantially more sample-efficient. We first show that MGT and HWT latent representations are linearly separable in low-dimensional space, and provide a plausible explanation for this separability through systematic differences in their representation quality. Motivated by these insights, we train two variants of simple linear probes and evaluate them across 4 benchmarks against 16 baselines. Probes consistently improve OOD detection (+11 AUC), requiring solely ${<}100$ samples to reach near-peak performance. We show that this transferability arises because probes recover a shared latent MGT direction that generalizes across diverse settings. Finally, we demonstrate that probing vectors capture a continuous spectrum of ``machineness'', highlighting their potential for fine-grained estimation of AI-edited text. Overall, our work provides insights into latent-space differences between MGT and HWT and demonstrates the potential of linear probes as as robust and sample-efficient MGT detectors. We release our code on~\href{ this https URL }{github}.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-26 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.24780 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T06:17:29.383Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-26?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).