Pith. sign in

REVIEW 3 major objections 5 minor 50 references

Mitigating Out-of-Entity Errors in Named Entity Recognition: A Sentence-Level Strategy

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A sentence-level context vector refined by template contrastive learning lifts OOE-NER F1 from 74.78 to 78.69 across five benchmarks.

desk verdict A plausible sentence-level context module for OOE-NER that needs a dev-set template selection before the headline numbers can be trusted. read the letter →

arxiv 2412.08434 v2 pith:7NI2IJ6R submitted 2024-12-11 cs.CL cs.AI

Guochao Jiang , Ziqin Luo , Chengwei Hu , Zepeng Ding , Deqing Yang This is my paper · ORCID
classification cs.CLcs.AI
keywords namedentityrecognitionout-of-entityOOE-NERsentencerepresentationcontrastivelearningtemplatepoolingspanclassificationGPT-4templates
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Out-of-entity (OOE) errors are the NER failures that occur when the tokens of a test entity mention never appeared in training, leaving the span's own representation uninformative. This paper claims that those errors can be substantially reduced by adding a representation of the whole sentence to the span representation and refining that sentence vector with a contrastive loss against positive and negative template sentences. The proposed framework, S+NER, computes a sentence-level vector $c$ by average-pooling BERT's token representations, concatenates it onto every candidate span, and trains it with templates such as '[SPAN] is a [TYPE] entity.' generated by GPT-4. On five OOE benchmarks, S+NER reports an average entity F1 of 78.69, compared with 74.78 for the strongest baseline, and its largest gain comes on WNUT2017, where 100% of test entities are out-of-entity. If the claim is right, it means that for never-seen mention words, the sentence's statement about what kind of thing the mention is can substitute for knowledge of the mention's own spelling.

What carries the argument

The central machinery is the sentence-level context vector $c$, defined in Eq. (5) as the average of all token representations produced by a BERT-based encoder. $c$ is concatenated onto every span representation in Eq. (6), then refined during training by the template contrastive loss in Eq. (7), an InfoNCE-style objective that aligns $c$ with pooled positive template representations (e.g., '[SPAN] is a [TYPE] entity.' filled with the correct type) and separates it from negative templates (wrong type or non-entity). Template pooling aggregates the multiple templates generated by GPT-4 and manually selected, and the refined $c$ is what lets the classifier exploit type-discriminative sentence evidence even when the mention's own tokens are novel.

What would settle it

Replace the average-pooled sentence vector $c$ from Eq. (5) with the [CLS] token embedding, keeping the template contrastive loss and all other hyperparameters fixed; if entity F1 on WNUT2017 (58.27), TwitterNER (78.01), and CoNLL2003-OOE (83.52) does not fall materially, average pooling is not doing the work the paper assigns to it.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that the OOE gap in span-based NER is largely recoverable from the sentence itself. S+NER takes a span classification backbone, computes $c$ by average-pooling all token representations from a BERT-based encoder (Eq. 5), and appends $c$ to each span's boundary and length embeddings before classification (Eq. 6). During training it refines $c$ with an InfoNCE-style contrastive loss (Eq. 7): the sentence vector is pulled toward the pooled representations of positive templates, which state the span with its correct type, and pushed away from negative templates, which state a wrong type or non-entity. The templates are generated by GPT-4, manually filtered to ten, and combined by template pooling. With this design, the model reports entity F1 gains on all five benchmarks, averaging 78.69 against 74.78 for the best baseline DSpERT, and the method works with BERT, RoBERTa, and DeBERTa encoders alike.

Load-bearing premise

The method assumes that averaging the token representations of a sentence yields a single vector that carries enough entity-type evidence for a never-seen mention, and that the hand-picked template sentences used to refine it are the right ones for every dataset.

Editorial extensions

If this is right

  • S+NER beats every compared baseline on WNUT2017, TwitterNER, CoNLL2003-Typos, and CoNLL2003-OOE, and the average entity F1 over the five datasets is 78.69 against 74.78 for DSpERT.
  • Ablations show both additions carry weight: on WNUT2017, SpanNER alone scores 51.83, adding the context representation raises it to 53.21, and adding template pooling plus contrastive learning raises it to 58.27.
  • The gain holds across different encoders: S+NER outperforms SpanNER with BERT-large, RoBERTa-large, and DeBERTa-large on all three tested datasets.
  • S+NER maintains its advantage over SpanNER, DSpERT, and MINER at every OOE rate from 50% to 100%, even though all models degrade as the OOE rate rises.
  • The framework is sensitive to template wording, but pooling over the ten templates chosen outperforms each single template, so the multi-template design is doing work.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A strict transfer test would freeze the ten templates selected on WNUT2017 and apply them to the other four datasets without retuning; the paper's template choice is made on WNUT2017 test performance, so this is a separate check of generality.
  • Because the same $c$ is shared by all spans in a sentence, a sentence containing multiple entities of different types forces one vector to move toward several positive templates at once; a testable prediction is that S+NER's per-entity gain shrinks as the number of distinct entity types in one sentence grows.
  • The contrastive refinement is not tied to span-based decoding: any encoder that produces token representations could emit $c$ and be trained with the same template loss, so sequence labeling and generative NER decoders are natural next testbeds for the same idea.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper addresses Out-of-Entity (OOE) named entity recognition, where test entity mentions contain tokens that were not seen during training. It proposes S+NER, a span-based NER framework that augments span representations with a sentence-level representation (Eqs. 5–6) and refines that representation via contrastive learning against positive and negative textual templates generated with GPT-4 (Eqs. 7–8, Section 3.2.2). The method is evaluated on five OOE-oriented datasets, with entity-level micro F1 as the main metric, and is compared against several baselines including SpanNER, MINER, and DSpERT (Table 2). The paper also reports ablations (Table 4), template sensitivity (Table 5), encoder variation (Table 6), and performance across different OOE rates (Figure 3).

Significance. The central idea—using sentence-level context as an auxiliary signal for unseen entity mentions and aligning it with textual templates—is reasonable and potentially useful for OOE-NER. The paper has a strong empirical scaffold: five datasets, ablations separating the contribution of the context representation from the contrastive template module, and encoder-agnostic checks with BERT, RoBERTa, and DeBERTa. If the reported numbers are unbiased, the improvements are practically meaningful, e.g., an average F1 of 78.69 versus 74.78 for DSpERT. However, the template selection protocol appears to use the WNUT2017 test set, which biases the headline result and the average comparison; additionally, no variance or significance tests are reported, so smaller margins such as the 1.24 F1 gain on JNLPBA are hard to interpret. The manuscript is a valuable contribution in progress, but the evaluation needs a selection-protocol fix and statistical rigor before the claims can be accepted.

major comments (3)
  1. [Section 3.2.2 / Table 5] The template set is selected using the WNUT2017 test set, which biases the headline result. Section 3.2.2 states that 100 GPT-4 templates were generated and that 10 representative templates were manually selected, while Section 4.5 reports per-template micro F1 scores on WNUT2017 (Table 5). If the manual selection is informed by those test-set scores, then the WNUT2017 result (58.27 vs. 55.32 for DSpERT) is an optimistic selection artifact, and because the same 10 templates are reused on the other four datasets, the average comparison is also affected. Since WNUT2017 has a 100% OOE rate, the paper's central robustness claim rests on this number. Please select templates using only development data (or a nested selection procedure), report the selected template set, and state the selection criterion explicitly.
  2. [Section 4.2 / Tables 2, 4, 5, 6] No variance or significance information is reported, despite Section 4.2 stating that F1 scores are averages over five random experiments. For example, the JNLPBA margin over DSpERT is 1.24 F1 (82.70 vs. 81.46), and the template-pooling gain over the best single template in Table 5 is 1.35 F1; without standard deviations or paired significance tests, these differences are not clearly distinguished from noise. Please report per-seed results and run paired significance tests (e.g., bootstrap or approximate randomization) across the five seeds.
  3. [Section 3.2.2 / Eq. (7)] The template-pooling operation is underspecified. Eq. (7) writes InfoNCE(c, Pooling(T_i^+), Pooling(T_i^-)) but does not define the pooling function (mean, max, or other), how the positive and negative template sets are assembled from the 10 templates, or which 10 templates were selected from the 100 generated by GPT-4. Since template pooling is a named contribution and the model is sensitive to template choice (Table 5), the missing definitions and template list prevent reproduction. Please state the pooling operation, list all selected templates, and include the GPT-4 prompt or generation procedure.
minor comments (5)
  1. [Section 1 / Figure 1] The text mentions 'MIENR' with the citation to Wang et al. (2022), but the cited model is MINER; please correct the name for consistency.
  2. [Section 4.3] The text says that InferNER, CoFEE, and MAML scores are taken from the MINER paper, but Table 2 does not include a MAML row; please either add the row or remove the mention.
  3. [Section 3.2.2 / Eq. (7)] Eq. (7) omits the temperature parameter τ that Table 3 lists as 1; including τ explicitly in the InfoNCE formula would improve clarity.
  4. [Section 4.2] The sentence 'To limit the number of all extracted spans on the affordable level' is awkward; consider rephrasing as 'To keep the number of extracted spans manageable.'
  5. [Table 5] Table 5 lists only four of the ten selected templates; please provide the full set of ten templates, perhaps in an appendix, so that the template-pooling comparison is reproducible.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the S+NER gains come from a train/test empirical pipeline, not from a derivation that reduces to its own inputs.

full rationale

The paper's central derivation is an empirical pipeline: Eq. (5) obtains a sentence representation by average-pooling BERT token representations, Eq. (6) concatenates that representation with the span boundary and length embeddings, Eq. (7) defines an InfoNCE-style contrastive loss against positive and negative template representations, and Eq. (8) combines it with the span classification loss. None of these equations defines the reported entity F1 gains in terms of the test labels, fitted constants, or the method's own outputs; the model is trained on training splits and evaluated on held-out test splits, with the checkpoint selected by validation performance as stated in Section 4.2. The self-citations to ToNER (Jiang et al., 2024b) and P-ICL (Jiang et al., 2024a) appear only in the introduction and related work, and no load-bearing claim or uniqueness theorem is imported from them. The template-selection protocol is a potential evaluation-validity concern worth noting: Section 3.2.2 says 10 representative templates were manually selected from 100 GPT-4-generated templates, and Section 4.5 reports per-template micro F1 scores on WNUT2017. If the manual selection had used those test-set scores, the WNUT2017 result would be optimistically biased, but the paper does not state that the selection was based on Table 5, and the same template set is reused across all five datasets. This is a benchmark-integrity or tuning issue, not a definitional or self-citation circularity. Since no circular step can be exhibited from the paper's text, the correct verdict is no significant circularity.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The ledger shows a small set of hyperparameters and the core template-related postulates. The paper does not introduce new physical or mathematical entities. The template set is a hand-crafted artifact that the central claim depends on, distinguishing it from standard hyperparameters.

free parameters (4)
  • lambda (contrastive loss weight) = 0.1
    Weight of the contrastive loss L2 in the total loss (Eq. 8); chosen by tuning studies.
  • tau (InfoNCE temperature) = 1
    Temperature hyperparameter in the contrastive loss; chosen by tuning.
  • span length embedding dimension d' = 50
    Dimension of the span length embedding, part of the span representation (Eq. 3).
  • manual template set = 10 templates selected from 100 GPT-4-generated
    The set of templates used for contrastive learning was manually selected based on performance on WNUT2017 (Section 4.5); this selection affects the final results.
assumptions (3)
  • domain assumption Contextual information is highly correlated with the correct understanding of the entity and its type.
    This is the motivating assumption in the Introduction and Section 3.2; the entire method depends on this correlation being strong enough to improve OOE-NER.
  • domain assumption Pre-trained language models encode sentence-level semantic and syntactic information useful for NER.
    The method uses BERT average-pooled sentence embeddings (Section 3.2.1, Eq. 5) and assumes these capture entity-type evidence.
  • ad hoc to paper GPT-4-generated templates, when filled with [SPAN] and [TYPE], serve as golden-standard representations for contrastive learning.
    Section 3.2.2 introduces the template set as the positive/negative anchors; this is the paper's central modeling postulate.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Mitigating Out-of-Entity Errors in Named Entity Recognition: A Sentence-Level Strategy." pith.science (2026). https://pith.science/paper/7NI2IJ6R

@misc{pith2026241208434,
  author       = {Pith},
  title        = {Pith review of: Mitigating Out-of-Entity Errors in Named Entity Recognition: A Sentence-Level Strategy},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7NI2IJ6R}},
  note         = {Machine review of arXiv:2412.08434}
}
read the original abstract

Many previous models of named entity recognition (NER) suffer from the problem of Out-of-Entity (OOE), i.e., the tokens in the entity mentions of the test samples have not appeared in the training samples, which hinders the achievement of satisfactory performance. To improve OOE-NER performance, in this paper, we propose a new framework, namely S+NER, which fully leverages sentence-level information. Our S+NER achieves better OOE-NER performance mainly due to the following two particular designs. 1) It first exploits the pre-trained language model's capability of understanding the target entity's sentence-level context with a template set. 2) Then, it refines the sentence-level representation based on the positive and negative templates, through a contrastive learning strategy and template pooling method, to obtain better NER results. Our extensive experiments on five benchmark datasets have demonstrated that, our S+NER outperforms some state-of-the-art OOE-NER models.

Figures

Figures reproduced from arXiv: 2412.08434 by the authors.

Figure 1
Figure 1. The span-based NER models’ F1 scores on TwitterNER dataset with different OOE rates. The OOE rate is defined as the ratio of the test entities whose men￾tion spans have the words (tokens) not appearing in the training set, to all entities in the test set. The numbers here are the results of deduplication, and duplicate enti￾ties are not considered. when many words (tokens) in the entity mentions have not appeared in… view at source ↗
Figure 2
Figure 2. The overall architecture of our proposed S+NER (better viewed in color), which has two major parts: [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. The performance of SpanNER, DSpERT, MINER and S+NER on different OOE rates of Twit￾terNER. It is worth noting that the TwitterNER dataset here is re-partitioned the training and test set to achieve different OOE rates. 3, where their performance on six different OOE rates are displayed. It shows that, although all models’ performance degrades as the OOE rate in￾creases, S+NER outperforms the three baselines on all O… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 20 canonical work pages

  1. [1]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774

    arXiv 2023

  2. [2]

    Alan Akbik, Duncan Blythe, and Roland Vollgraf. 2018. https://aclanthology.org/C18-1139/ Contextual string embeddings for sequence labeling . In Proceedings of the 27th International Conference on Computational Linguistics, COLING 2018, Santa Fe, New Mexico, USA, August 20-26, 2018 , pages 1638--1649. Association for Computational Linguistics

    work page 2018

  3. [3]

    Diego Moll \' a Aliod, Menno van Zaanen, and Daniel Smith. 2006. https://aclanthology.org/U06-1009/ Named entity recognition for question answering . In Proceedings of the Australasian Language Technology Workshop, ALTA 2006, Sydney, Australia, November 30-December 1, 2006 , pages 51--58. Australasian Language Technology Association

    work page 2006

  4. [4]

    Partha Sarathy Banerjee, Baisakhi Chakraborty, Deepak Tripathi, Hardik Gupta, and Sourabh S. Kumar. 2019. https://doi.org/10.1007/S11277-019-06501-Z A information retrieval based on question and answering and NER for unstructured information without using SQL . Wirel. Pers. Commun., 108(3):1909--1931

    work page doi:10.1007/s11277-019-06501-z 2019

  5. [6]

    Jason P. C. Chiu and Eric Nichols. 2016. https://doi.org/10.1162/TACL\_A\_00104 Named entity recognition with bidirectional lstm-cnns . Trans. Assoc. Comput. Linguistics, 4:357--370

    doi:10.1162/tacl 2016

  6. [7]

    Nigel Collier and Jin - Dong Kim. 2004. https://aclanthology.org/W04-1213/ Introduction to the bio-entity recognition task at JNLPBA . In Proceedings of the International Joint Workshop on Natural Language Processing in Biomedicine and its Applications, NLPBA/BioNLP 2004, Geneva, Switzerland, August 28-29, 2004

    work page 2004

  7. [8]

    Leyang Cui, Yu Wu, Jian Liu, Sen Yang, and Yue Zhang. 2021. https://doi.org/10.18653/V1/2021.FINDINGS-ACL.161 Template-based named entity recognition using BART . In Findings of the Association for Computational Linguistics: ACL/IJCNLP 2021, Online Event, August 1-6, 2021 , volume ACL/IJCNLP 2021 of Findings of ACL , pages 1835--1845. Association for Comp...

    doi:10.18653/v1/2021.findings-acl.161 2021

  8. [9]

    Xiang Dai and Heike Adel. 2020. https://doi.org/10.18653/V1/2020.COLING-MAIN.343 An analysis of simple data augmentation for named entity recognition . In Proceedings of the 28th International Conference on Computational Linguistics, COLING 2020, Barcelona, Spain (Online), December 8-13, 2020 , pages 3861--3867. International Committee on Computational Li...

    doi:10.18653/v1/2020.coling-main.343 2020

Show all 50 references
  1. [10]

    Passonneau, and Rui Zhang

    Sarkar Snigdha Sarathi Das, Arzoo Katiyar, Rebecca J. Passonneau, and Rui Zhang. 2022. https://doi.org/10.18653/V1/2022.ACL-LONG.439 Container: Few-shot named entity recognition via contrastive learning . In Proceedings of the 60th Annual Meeting of the Association for Computa...

    2022 doi

  2. [11]

    Leon Derczynski, Eric Nichols, Marieke van Erp, and Nut Limsopatham. 2017. https://doi.org/10.18653/V1/W17-4418 Results of the WNUT2017 shared task on novel and emerging entity recognition . In Proceedings of the 3rd Workshop on Noisy User-generated Text, NUT@EMNLP 2017, Copen...

    2017 doi

  3. [12]

    Jacob Devlin, Ming - Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. https://doi.org/10.18653/V1/N19-1423 BERT: pre-training of deep bidirectional transformers for language understanding . In Proceedings of the 2019 Conference of the North American Chapter of the Associat...

    2019 doi

  4. [13]

    Jinlan Fu, Xuanjing Huang, and Pengfei Liu. 2021. https://doi.org/10.18653/V1/2021.ACL-LONG.558 Spanner: Named entity re-/recognition as span prediction . In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Join...

    2021 doi

  5. [14]

    Nobukazu Fukuda, Naoki Yoshinaga, and Masaru Kitsuregawa. 2020. https://doi.org/10.18653/V1/2020.FINDINGS-EMNLP.434 Robust backed-off estimation of out-of-vocabulary embeddings . In Findings of the Association for Computational Linguistics: EMNLP 2020, Online Event, 16-20 Nove...

    2020 doi

  6. [15]

    Jonas Gehring, Michael Auli, David Grangier, Denis Yarats, and Yann N. Dauphin. 2017. http://proceedings.mlr.press/v70/gehring17a.html Convolutional sequence to sequence learning . In Proceedings of the 34th International Conference on Machine Learning, ICML 2017, Sydney, NSW,...

    2017

  7. [16]

    Pengcheng He, Xiaodong Liu, Jianfeng Gao, and Weizhu Chen. 2021. https://openreview.net/forum?id=XPZIaotutsD Deberta: decoding-enhanced bert with disentangled attention . In 9th International Conference on Learning Representations, ICLR 2021, Virtual Event, Austria, May 3-7, 2...

    2021

  8. [17]

    Sepp Hochreiter and J \" u rgen Schmidhuber. 1997. https://doi.org/10.1162/NECO.1997.9.8.1735 Long short-term memory . Neural Comput., 9(8):1735--1780

    1997 doi

  9. [18]

    Ziniu Hu, Ting Chen, Kai - Wei Chang, and Yizhou Sun. 2019. https://doi.org/10.18653/V1/P19-1402 Few-shot representation learning for out-of-vocabulary words . In Proceedings of the 57th Conference of the Association for Computational Linguistics, ACL 2019, Florence, Italy, Ju...

    2019 doi

  10. [19]

    Guochao Jiang, Zepeng Ding, Yuchen Shi, and Deqing Yang. 2024 a . P-icl: Point in-context learning for named entity recognition with large language models. arXiv preprint arXiv:2405.04960

    2024 arXiv

  11. [20]

    Guochao Jiang, Ziqin Luo, Yuchen Shi, Dixuan Wang, Jiaqing Liang, and Deqing Yang. 2024 b . https://aclanthology.org/2024.lrec-main.1412 Toner: Type-oriented named entity recognition with generative language model . In Proceedings of the 2024 Joint International Conference on ...

    2024

  12. [21]

    Changliang Li, Liang Li, and Ji Qi. 2018. https://doi.org/10.18653/V1/D18-1417 A self-attentive model with gate mechanism for spoken language understanding . In Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing, Brussels, Belgium, October 3...

    2018 doi

  13. [22]

    Jing Li, Aixin Sun, Jianglei Han, and Chenliang Li. 2022. https://doi.org/10.1109/TKDE.2020.2981314 A survey on deep learning for named entity recognition . IEEE Trans. Knowl. Data Eng. , 34(1):50--70

    2022

  14. [23]

    Yinhan Liu. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692

    2019 arXiv

  15. [24]

    Xuezhe Ma and Eduard H. Hovy. 2016. https://doi.org/10.18653/V1/P16-1101 End-to-end sequence labeling via bi-directional lstm-cnns-crf . In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics, ACL 2016, August 7-12, 2016, Berlin, Germany, Vo...

    2016 doi

  16. [25]

    Andrei Mikheev, Marc Moens, and Claire Grover. 1999. https://aclanthology.org/E99-1001/ Named entity recognition without gazetteers . In EACL 1999, 9th Conference of the European Chapter of the Association for Computational Linguistics, June 8-12, 1999, University of Bergen, B...

    1999

  17. [26]

    Aaron van den Oord, Yazhe Li, and Oriol Vinyals. 2018. Representation learning with contrastive predictive coding. arXiv preprint arXiv:1807.03748

    2018 arXiv

  18. [27]

    Minlong Peng, Qi Zhang, Xiaoyu Xing, Tao Gui, Jinlan Fu, and Xuanjing Huang. 2019. https://doi.org/10.24963/IJCAI.2019/715 Learning task-specific representation for novel words in sequence labeling . In Proceedings of the Twenty-Eighth International Joint Conference on Artific...

    2019 doi

  19. [28]

    Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer

    Matthew E. Peters, Mark Neumann, Mohit Iyyer, Matt Gardner, Christopher Clark, Kenton Lee, and Luke Zettlemoyer. 2018. https://doi.org/10.18653/V1/N18-1202 Deep contextualized word representations . In Proceedings of the 2018 Conference of the North American Chapter of the Ass...

    2018 doi

  20. [29]

    Yuval Pinter, Robert Guthrie, and Jacob Eisenstein. 2017. https://doi.org/10.18653/V1/D17-1010 Mimicking word embeddings using subword rnns . In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing, EMNLP 2017, Copenhagen, Denmark, September 9...

    2017 doi

  21. [30]

    Yujia Qin, Yankai Lin, Ryuichi Takanobu, Zhiyuan Liu, Peng Li, Heng Ji, Minlie Huang, Maosong Sun, and Jie Zhou. 2021. https://doi.org/10.18653/V1/2021.ACL-LONG.260 ERICA: improving entity and relation understanding for pre-trained language models via contrastive learning . In...

    2021 doi

  22. [31]

    Moemmur Shahzad, Ayesha Amin, Diego Esteves, and Axel - Cyrille Ngonga Ngomo. 2021. https://doi.org/10.32473/FLAIRS.V34I1.128538 Inferner: an attentive model leveraging the sentence-level information for named entity recognition in microblogs . In Proceedings of the Thirty-Fou...

    2021 doi

  23. [32]

    Yuchen Shi, Guochao Jiang, Tian Qiu, and Deqing Yang. 2024. https://doi.org/10.1145/3627673.3679791 Agentre: An agent-based framework for navigating complex information landscapes in relation extraction . In Proceedings of the 33rd ACM International Conference on Information a...

    2024

  24. [33]

    Jianlin Su, Murtadha H. M. Ahmed, Yu Lu, Shengfeng Pan, Wen Bo, and Yunfeng Liu. 2024. https://doi.org/10.1016/J.NEUCOM.2023.127063 Roformer: Enhanced transformer with rotary position embedding . Neurocomputing, 568:127063

    2024

  25. [34]

    Gomez, Lukasz Kaiser, and Illia Polosukhin

    Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Lukasz Kaiser, and Illia Polosukhin. 2017. https://proceedings.neurips.cc/paper/2017/hash/3f5ee243547dee91fbd053c1c4a845aa-Abstract.html Attention is all you need . In Advances in Neural I...

    2017

  26. [35]

    Jiuniu Wang, Wenjia Xu, Xingyu Fu, Guangluan Xu, and Yirong Wu. 2020. https://doi.org/10.1016/J.KNOSYS.2020.105842 ASTRAL: adversarial trained LSTM-CNN for named entity recognition . Knowl. Based Syst., 197:105842

    2020

  27. [36]

    Tongzhou Wang and Phillip Isola. 2020. http://proceedings.mlr.press/v119/wang20k.html Understanding contrastive representation learning through alignment and uniformity on the hypersphere . In Proceedings of the 37th International Conference on Machine Learning, ICML 2020, 13-...

    2020

  28. [37]

    Xiao Wang, Shihan Dou, Limao Xiong, Yicheng Zou, Qi Zhang, Tao Gui, Liang Qiao, Zhanzhan Cheng, and Xuanjing Huang. 2022. https://doi.org/10.18653/V1/2022.ACL-LONG.383 MINER: improving out-of-vocabulary named entity recognition from an information theoretic perspective . In Pr...

    2022 doi

  29. [38]

    Xiao Wang, Qin Liu, Tao Gui, Qi Zhang, Yicheng Zou, Xin Zhou, Jiacheng Ye, Yongxin Zhang, Rui Zheng, Zexiong Pang, Qinzhuo Wu, Zhengyan Li, Chong Zhang, Ruotian Ma, Zichu Fei, Ruijian Cai, Jun Zhao, Xingwu Hu, Zhiheng Yan, Yiding Tan, Yuan Hu, Qiyuan Bian, Zhihua Liu, Shan Qin...

    2021 doi

  30. [39]

    Bo Xu, Yong Xu, Jiaqing Liang, Chenhao Xie, Bin Liang, Wanyun Cui, and Yanghua Xiao. 2017. https://doi.org/10.1007/978-3-319-60045-1\_44 Cn-dbpedia: A never-ending chinese knowledge extraction system . In Advances in Artificial Intelligence: From Theory to Practice - 30th Inte...

    2017 doi

  31. [40]

    Mengge Xue, Bowen Yu, Zhenyu Zhang, Tingwen Liu, Yue Zhang, and Bin Wang. 2020. https://doi.org/10.18653/V1/2020.EMNLP-MAIN.514 Coarse-to-fine pre-training for named entity recognition . In Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing,...

    2020 doi

  32. [41]

    Ikuya Yamada, Akari Asai, Hiroyuki Shindo, Hideaki Takeda, and Yuji Matsumoto. 2020. https://doi.org/10.18653/V1/2020.EMNLP-MAIN.523 LUKE: deep contextualized entity representations with entity-aware self-attention . In Proceedings of the 2020 Conference on Empirical Methods i...

    2020 doi

  33. [42]

    Hang Yan, Bocao Deng, Xiaonan Li, and Xipeng Qiu. 2019. Tener: adapting transformer encoder for named entity recognition. arXiv preprint arXiv:1911.04474

    2019 arXiv

  34. [43]

    Hang Yan, Tao Gui, Junqi Dai, Qipeng Guo, Zheng Zhang, and Xipeng Qiu. 2021. https://doi.org/10.18653/V1/2021.ACL-LONG.451 A unified generative framework for various NER subtasks . In Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and t...

    2021 doi

  35. [44]

    Qi Zhang, Jinlan Fu, Xiaoyu Liu, and Xuanjing Huang. 2018. https://doi.org/10.1609/AAAI.V32I1.11962 Adaptive co-attention network for named entity recognition in tweets . In Proceedings of the Thirty-Second AAAI Conference on Artificial Intelligence, (AAAI-18), the 30th innova...

    2018 doi

  36. [45]

    Sheng Zhang, Hao Cheng, Jianfeng Gao, and Hoifung Poon. 2023. https://openreview.net/forum?id=9EAQVEINuum Optimizing bi-encoder for named entity recognition via contrastive learning . In The Eleventh International Conference on Learning Representations, ICLR 2023, Kigali, Rwan...

    2023

  37. [46]

    Yafeng Zhang, Zilan Yu, Yuang Huang, and Jing Tang. 2024. Cllmfs: A contrastive learning enhanced large language model framework for few-shot named entity recognition. arXiv preprint arXiv:2408.12834

    2024 arXiv

  38. [47]

    Yue Zhang and Jie Yang. 2018. https://doi.org/10.18653/V1/P18-1144 Chinese NER using lattice LSTM . In Proceedings of the 56th Annual Meeting of the Association for Computational Linguistics, ACL 2018, Melbourne, Australia, July 15-20, 2018, Volume 1: Long Papers , pages 1554-...

    2018 doi

  39. [48]

    Enwei Zhu, Yiyang Liu, and Jinpeng Li. 2023. https://doi.org/10.18653/V1/2023.FINDINGS-ACL.672 Deep span representations for named entity recognition . In Findings of the Association for Computational Linguistics: ACL 2023, Toronto, Canada, July 9-14, 2023 , pages 10565--10582...

    2023 doi

  40. [49]

    Qile Zhu, Xiaolin Li, Ana Conesa, and C \' e cile Pereira. 2018. https://doi.org/10.1093/BIOINFORMATICS/BTX815 GRAM-CNN: a deep learning approach with local context for named entity recognition in biomedical text . Bioinform., 34(9):1547--1554

    2018 doi

  41. [50]

    online" 'onlinestring :=

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint eprinttype howpublished institution journal key month note number organization pages publisher school series title type volume year doi pubmed url lastchecked label extra.label sort.label short.list...

  42. [51]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.