Metraipolis › archive › dc_2138b45eaaae

WiC is Not WSD: A Study on LLMs and Lexical Ambiguity Resolution — arXiv:2609.20593

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.20593
License (as archived)
CC0-1.0
Archived
2026-10-10T03:48:24.672Z
SHA-256 of the archived text
e0b67789109ec5674ca10a6c709ae071b7a26eb65d2b74c82eebe79bd88efefc
Stable URL of this page
https://metraipolis.com/documento/dc_2138b45eaaae

How to cite

WiC is Not WSD: A Study on LLMs and Lexical Ambiguity Resolution — arXiv:2609.20593 — arXiv (en), https://arxiv.org/abs/2609.20593 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_2138b45eaaae (Metraipolis archive dc_2138b45eaaae, sha256 e0b67789109ec5674ca10a6c709ae071b7a26eb65d2b74c82eebe79bd88efefc)
BibTeX
@misc{dc2138b45eaaae,
  title = {WiC is Not WSD: A Study on LLMs and Lexical Ambiguity Resolution — arXiv:2609.20593},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.20593},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_2138b45eaaae; sha256 e0b67789109ec5674ca10a6c709ae071b7a26eb65d2b74c82eebe79bd88efefc},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.20593. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Word-in-Context (WiC) remains challenging for language models, despite recent progress on lexical-semantic tasks. We hypothesise that this difficulty arises not only from comparing two contextual uses of a word, but also from the absence of an explicit sense inventory that specifies the relevant level of semantic granularity. We evaluate open LLMs on WiC and traditional Word Sense Disambiguation (WSD) under similar settings. We find that providing candidate senses, similar to what is done in traditional WSD, improves WiC performance in all settings. In general, explicit sense information helps models make more consistent and targeted judgements. Human evaluation further shows that many apparent WiC errors reflect label ambiguity or mismatches between model and annotator sense boundaries rather than simple failures of lexical understanding. In particular, results show that LLMs overthink the sense distinction often leading to errors based on overly fine-grained distinctions.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-18 in cs.CL).
Categorie arXiv: Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2609.20593 — 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 2026-10-10T03:24:08.167Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).