Metraipolis › archive › dc_08249fa41c8e

Not All Irregularity Is Equal: Causally Isolating a Rare Failure Mode in Japanese Morphological Inflection — arXiv:2609.21179

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.21179
License (as archived)
CC0-1.0
Archived
2026-10-10T01:27:14.645Z
SHA-256 of the archived text
1e2b2222fbe8280d796dbdb6bab00f2878f86ffb94e2695cddfb92966196a4dd
Stable URL of this page
https://metraipolis.com/documento/dc_08249fa41c8e

How to cite

Not All Irregularity Is Equal: Causally Isolating a Rare Failure Mode in Japanese Morphological Inflection — arXiv:2609.21179 — arXiv (en), https://arxiv.org/abs/2609.21179 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_08249fa41c8e (Metraipolis archive dc_08249fa41c8e, sha256 1e2b2222fbe8280d796dbdb6bab00f2878f86ffb94e2695cddfb92966196a4dd)
BibTeX
@misc{dc08249fa41c8e,
  title = {Not All Irregularity Is Equal: Causally Isolating a Rare Failure Mode in Japanese Morphological Inflection — arXiv:2609.21179},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.21179},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_08249fa41c8e; sha256 1e2b2222fbe8280d796dbdb6bab00f2878f86ffb94e2695cddfb92966196a4dd},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.21179. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Neural morphological generation systems often achieve high aggregate accuracy on benchmark datasets, yet such performance can conceal systematic errors clustered in rare morphological subclasses. We present an orthography-aware diagnosis of Japanese past-tense verb inflection, treating hiragana not merely as a transcriptional medium but as a representational system that encodes morphophonological structure. Using two character-level Transformer architectures evaluated across five random seeds, we show that although both systems exceed 97% aggregate accuracy, a single structurally specific irregular subtype, verbs whose stems end in /e/ and require gemination before the past-tense suffix and make up fewer than 1% of the data, accounts for a disproportionate 30-43% share of residual errors and contributes roughly 34-48x its prevalence to total errors. We then move from diagnosis to causal isolation: controlled ablation experiments show that removing this subtype alone produces larger accuracy gains than removing all irregular verbs combined. These findings indicate that error concentration in neural morphological learning is not driven by irregularity per se, but by the interaction between extreme low-frequency morphological patterns and specific orthographic processes. We argue that morphological evaluation should incorporate fine-grained subclass analysis, and discuss implications for data-efficient, developmentally plausible language model pretraining.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.21179 — 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 2026-10-10T01:00:30.269Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).