@inproceedings{morand-etal-2026-tommer,
    title = "{T}o{MM}e{R} - Efficient Entity Mention Detection from Large Language Models",
    author = "Morand, Victor  and
      Tomeh, Nadi  and
      Mothe, Josiane  and
      Piwowarski, Benjamin",
    editor = "Liakata, Maria  and
      Moreira, Viviane P.  and
      Zhang, Jiajun  and
      Jurgens, David",
    booktitle = "Proceedings of the 64th Annual Meeting of the {A}ssociation for {C}omputational {L}inguistics (Volume 1: Long Papers)",
    month = jul,
    year = "2026",
    address = "San Diego, California, United States",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-acl/2026.acl-long.1268/",
    pages = "27489--27509",
    ISBN = "979-8-89176-390-6",
    abstract = "Identifying which text spans refer to entities - mention detection- is both foundational for information extraction and a known performance bottleneck. We introduce ToMMeR, a lightweight model ({\ensuremath{<}}300K parameters) probing mention detection capabilities from early LLM layers. Across 13 NER benchmarks, ToMMeR achieves 93{\%} recall zero-shot, with an estimated 90{\%} precision under a human-calibrated LLM-judge protocol, showing that ToMMeR rarely produces spurious predictions despite high recall. Cross-model analysis reveals that diverse architectures (14M-15B parameters) converge on similar mention boundaries (DICE {\ensuremath{>}}75{\%}), confirming that mention detection emerges naturally from language modeling. When extended with span classification heads, ToMMeR achieves competitive NER performance (80-87{\%} F1 on standard benchmarks). Our work provides evidence that structured entity representations exist in early transformer layers and can be efficiently recovered with minimal parameters."
}