Metraipolis › archive › dc_5beeba89da08

A Systematic Analysis of the Predictive Power of LM Surprisal in Reading Chinese — arXiv:2610.04898

Source: arXiv
Original: https://arxiv.org/abs/2610.04898
License (as archived): CC0-1.0
Archived: 2026-10-06T05:26:29.650Z
SHA-256 of the archived text: 14d686c26bb4360787bc86da92387f0ce922f829eeae1268cace28b7498ca709
Stable URL of this page: https://metraipolis.com/documento/dc_5beeba89da08

How to cite

A Systematic Analysis of the Predictive Power of LM Surprisal in Reading Chinese — arXiv:2610.04898 — arXiv, https://arxiv.org/abs/2610.04898 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_5beeba89da08 (Metraipolis archive dc_5beeba89da08, sha256 14d686c26bb4360787bc86da92387f0ce922f829eeae1268cace28b7498ca709)

BibTeX

@misc{dc5beeba89da08,
  title = {A Systematic Analysis of the Predictive Power of LM Surprisal in Reading Chinese — arXiv:2610.04898},
  howpublished = {arXiv},
  url = {https://arxiv.org/abs/2610.04898},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_5beeba89da08; sha256 14d686c26bb4360787bc86da92387f0ce922f829eeae1268cace28b7498ca709},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv, https://arxiv.org/abs/2610.04898. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

This study analyzes the predictive power of LM-derived, token-level surprisal on Mandarin Chinese reading times. We first propose the Shortest Matching Sequence (SMS), an alignment scheme that maps between the word segmentation assumed by eye-tracking corpora and the LMs' subword tokenization, as the two tokenizations often disagree in the context of Mandarin Chinese. Then, using a suite of Chinese-Pythia models (14M-1.4B) trained on scratch with 30B tokens, we examine how well surprisal predicts first fixation duration, gaze duration, and total reading time in three paragraph-level eye-tracking corpora of Mandarin Chinese (GECO-CN, HKP, and MECO). Contrary to previous null findings, our results show that surprisal is predictive of Chinese reading times. However, whether predictive power scales with model size and the amount of training is corpus-specific: bigger models predict better in GECO-CN, whereas inverse scaling emerges in HKP and, at the largest sizes, in MECO. Subsequently, we tested one possible explanation for the inverse scaling in HKP and found that checkpoints whose surprisal remains closer to $n$-gram statistics are better predictors of reading. All in all, the predictive power of surprisal on Chinese reading time measurements is corpus-specific, which cautions against drawing scaling conclusions from a single corpus.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2610.04898 — 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 2026-10-06T05:24:05.927Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).