Research
ToMMeR -- Efficient Entity Mention Detection from Large Language Models
arXiv:2510.19410v2 Announce Type: replace Abstract: Identifying which text spans refer to entities - mention detection - is both foundational for information extraction and a known performance bottlen
arXiv:2510.19410v2 Announce Type: replace 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 (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.
Related
- Just Pass Twice: Efficient Token Classification with LLMs for Zero-Shot NER
- Faithfulness-Aware Uncertainty Quantification for Fact-Checking the Output of Retrieval Augmented Generation
- Attribution, Citation, and Quotation: A Survey of Evidence-based Text Generation with Large Language Models
- RAGognizer: Hallucination-Aware Fine-Tuning via Detection Head Integration
Source: arXiv cs.CL | 2026-04-21