Pith. sign in

REVIEW 2 major objections 2 minor 40 references

Distilled small language models reach 0.89 precision and 0.88 recall for clinical note de-identification on standard hardware.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.3

2026-07-03 00:08 UTC pith:SKKEXQCK

load-bearing objection SHIELD adds a diverse new de-id dataset and shows a distilled local model can hit usable 0.89/0.88 micro scores without cloud calls. the 2 major comments →

arxiv 2605.03301 v2 pith:SKKEXQCK submitted 2026-05-05 cs.CL cs.AI

SHIELD: A Diverse Clinical Note Dataset and Distilled Small Language Models for Enterprise-Scale De-identification

Jose D. Posada , David Love , Somalee Datta , Priya Desai This is my paper
classification cs.CL cs.AI
keywords de-identificationclinical noteslanguage model distillationPHI extractionelectronic health recordson-premise deploymentdiversity samplingsmall language models
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The paper introduces the SHIELD dataset of 1,381 diverse clinical notes annotated with 10,229 gold-standard PHI spans to overcome the limitations of older benchmarks like i2b2. It first benchmarks large language models on this data to set a performance ceiling, then applies teacher-student distillation to transfer the capabilities into small models that run locally. These distilled models deliver competitive span-level performance while satisfying hospital data governance rules that prohibit sending PHI to cloud services. The work also reports cross-dataset results showing reliable transfer on structured PHI but persistent gaps on institution-specific categories.

Core claim

By building SHIELD with set-cover diversity sampling across demographic and document-type strata plus human-in-the-loop adjudication, and then distilling from large LLMs, small models can be produced that achieve micro-averaged span-level precision of 0.89 and recall of 0.88 for nine PHI categories while running on standard workstation hardware, enabling accurate de-identification entirely behind institutional firewalls.

What carries the argument

Teacher-student distillation framework that transfers de-identification capabilities from large language models to small language models trained on the diversity-sampled SHIELD dataset.

Load-bearing premise

The set-cover diversity sampling across demographic and document-type strata together with human-in-the-loop adjudication produces a dataset whose semantic and demographic coverage is sufficient to support the generalization and cross-dataset claims made for the distilled models.

What would settle it

Testing the distilled model on a fresh set of clinical notes from an institution whose demographic and document-type profile was not included in the original set-cover sampling and observing whether macro-averaged recall on universal PHI categories falls substantially below 0.81.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Cross-dataset evaluation shows that diversity-trained models generalize well on universal structured PHI categories.
  • Institution-specific entities remain hard to transfer in both directions between datasets.
  • Broad-coverage models should be paired with specialized models for high-volume, semi-structured note types.
  • The distilled models provide a cost-effective, locally deployable alternative to cloud APIs for enterprise-scale use.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Real-world hospital pipelines could combine one general distilled model with a small set of institution-tuned models to cover both common and rare note formats.
  • The same set-cover sampling approach used to build SHIELD could be reused to create training data for other clinical NLP tasks that currently lack demographic diversity.
  • Widespread on-premise deployment of such models would lower barriers to secondary EHR research in smaller or privacy-sensitive health systems.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 2 minor

Summary. The paper introduces the SHIELD dataset of 1,381 clinical notes containing 10,229 gold-standard PHI spans across 9 categories, constructed via set-cover diversity sampling over demographic and document-type strata plus human-in-the-loop adjudication. It benchmarks four LLMs to establish an upper bound, then applies teacher-student distillation to produce small, locally deployable models; the best distilled model reports micro-averaged span-level P=0.89 and R=0.88 on workstation hardware, with macro recall of 0.81 versus the teacher's 0.90. Cross-dataset experiments indicate strong transfer on universal structured PHI but poor transfer on institution-specific entities, motivating hybrid broad-plus-specialized pipelines. The dataset and a distilled DeBERTa-v3 model are released publicly.

Significance. If the reported metrics and generalization patterns are reproducible, the work supplies a timely, demographically diverse public resource that directly addresses the obsolescence of i2b2-era benchmarks and the governance barriers to cloud-based de-identification. The explicit release of both data and a runnable small model is a concrete strength that enables immediate enterprise deployment and follow-on research.

major comments (2)
  1. [Abstract, §3] Abstract and §3 (Methods): the central performance claims (micro P/R 0.89/0.88, macro recall gap of 0.09) rest on gold-standard annotations whose reliability is not quantified in the provided abstract; inter-annotator agreement, adjudication protocol details, train/dev/test split statistics, and any statistical significance tests must be reported explicitly in the methods to substantiate that the numbers support the generalization and distillation conclusions.
  2. [§5] §5 (Cross-dataset evaluation): the claim that diversity-trained models 'generalize well on universal structured PHI categories' while institution-specific entities 'remain hard to transfer in both directions' is load-bearing for the hybrid-model recommendation, yet the section supplies no per-category confusion matrices, error analysis, or quantitative comparison of transfer gaps; without these the distinction between universal and institution-specific transfer cannot be assessed.
minor comments (2)
  1. [Table 1] Table 1 (dataset statistics): the 9 PHI categories are listed but the per-category span counts and demographic stratification achieved by the set-cover sampler are not shown; adding these rows would make the diversity claim verifiable.
  2. [§4.2] §4.2 (distillation): the teacher-student temperature and loss weighting are not stated; these hyperparameters should be reported so the 0.81 macro-recall result can be reproduced.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address the two major comments point-by-point below and will revise the manuscript to incorporate the requested details.

read point-by-point responses

  1. Referee: [Abstract, §3] Abstract and §3 (Methods): the central performance claims (micro P/R 0.89/0.88, macro recall gap of 0.09) rest on gold-standard annotations whose reliability is not quantified in the provided abstract; inter-annotator agreement, adjudication protocol details, train/dev/test split statistics, and any statistical significance tests must be reported explicitly in the methods to substantiate that the numbers support the generalization and distillation conclusions.

    Authors: We agree that explicit quantification of annotation reliability is necessary to support the performance claims. The current §3 describes the human-in-the-loop adjudication but does not report inter-annotator agreement, split statistics, or significance tests. In the revision we will add Fleiss' kappa for the adjudication step, exact train/dev/test split sizes with stratification details, and McNemar's tests comparing distilled-model vs. teacher performance. These will be placed in §3 and briefly referenced from the abstract. revision: yes

  2. Referee: [§5] §5 (Cross-dataset evaluation): the claim that diversity-trained models 'generalize well on universal structured PHI categories' while institution-specific entities 'remain hard to transfer in both directions' is load-bearing for the hybrid-model recommendation, yet the section supplies no per-category confusion matrices, error analysis, or quantitative comparison of transfer gaps; without these the distinction between universal and institution-specific transfer cannot be assessed.

    Authors: We concur that the current §5 lacks the granularity needed to fully substantiate the universal vs. institution-specific distinction. We will add per-category recall tables for both directions of transfer, a dedicated error-analysis subsection with examples of failure modes, and quantitative gap metrics (e.g., per-category Δ-recall between SHIELD-trained and cross-dataset models). These additions will directly support the hybrid-pipeline recommendation. revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

This is an empirical dataset-and-model paper with no equations, fitted parameters, or self-referential derivations. Performance metrics are direct evaluation outcomes on held-out and cross-dataset splits. The set-cover sampling and human adjudication are operational methods whose results are measured rather than defined by construction. No load-bearing self-citations, uniqueness theorems, or ansatzes are present. The central claims rest on external benchmarks and reported measurements.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central performance claims rest on two domain assumptions about data quality and representativeness that are not independently verified in the abstract; no free parameters or invented entities are described.

axioms (2)
  • domain assumption Human-in-the-loop adjudication produces accurate gold-standard PHI spans across the 10,229 annotated instances
    All downstream model evaluations and distillation results are measured against these spans.
  • domain assumption Set-cover diversity sampling across demographic and document-type strata yields a dataset representative of modern clinical narratives
    This premise is invoked to justify why SHIELD overcomes the limitations of older benchmarks such as i2b2.

pith-pipeline@v0.9.1-grok · 5864 in / 1568 out tokens · 47653 ms · 2026-07-03T00:08:11.653695+00:00 · methodology

0 comments
read the original abstract

De-identification of clinical text is a prerequisite for the secondary use of electronic health records. Existing public benchmarks such as the i2b2 2006 and 2014 corpora are over a decade old and lack the semantic and demographic diversity of modern clinical narratives. Large Language Models (LLMs) reach state-of-the-art zero-shot extraction, but their use at enterprise scale is limited by computational cost and by hospital data governance that restricts sending Protected Health Information (PHI) to cloud APIs. We introduce SHIELD (Synthetic Human-annotated Identifier-replaced Entries for Learning and De-identification), a diverse clinical note dataset of 1,381 notes with 10,229 gold-standard PHI spans across 9 categories, built with set-cover diversity sampling across demographic and document-type strata and human-in-the-loop adjudication. We evaluate four LLMs (two proprietary, two open-weight) to establish a performance ceiling on SHIELD, then show that a teacher-student distillation framework transfers these capabilities into locally deployable Small Language Models. Our best distilled model reaches micro-averaged span-level precision of 0.89 and recall of 0.88 while running on standard workstation hardware. It trails its cloud teacher on per-category recall (0.90 vs. 0.81 macro-averaged) but remains competitive given its lower cost and on-premise deployability. Cross-dataset evaluation shows that diversity-trained models generalize well on universal structured PHI categories, while institution-specific entities remain hard to transfer in both directions, which suggests pairing broad-coverage models with specialized models for high-volume, semi-structured note types. We publicly release the SHIELD dataset and the distilled DeBERTa v3 model to provide an accurate, cost-effective de-identification pipeline deployable entirely behind institutional firewalls.

Figures

Figures reproduced from arXiv: 2605.03301 by David Love, Jose D. Posada, Priya Desai, Somalee Datta.

Figure 1
Figure 1. Figure 1: SHIELD annotation pipeline. Clinical notes are sampled from STARR-OMOP via diversity view at source ↗
Figure 2
Figure 2. Figure 2: Teacher-student distillation process. A small labeled sample is used to create prompts view at source ↗
Figure 3
Figure 3. Figure 3: Dataset overview. (a) Statistics for the three evaluation corpora. (b) PHI category distribu view at source ↗
Figure 8
Figure 8. Figure 8: 7 view at source ↗
Figure 4
Figure 4. Figure 4: Corpus divergence analysis. (a) Fréchet Text Distance decomposed into mean shift and view at source ↗
Figure 5
Figure 5. Figure 5: LLM benchmark on SHIELD. Span-level radar profiles showing precision (left) and recall view at source ↗
Figure 6
Figure 6. Figure 6: Span-level radar comparison of two distilled student models (DeBERTa v3, BioModern) view at source ↗
Figure 7
Figure 7. Figure 7: Span-level distillation comparison on SHIELD: Gemini 2.5 Flash (Teacher) vs. DeBERTa view at source ↗
Figure 8
Figure 8. Figure 8: Span-level precision (left) and recall (right) with bootstrap 95% CIs for four transformer view at source ↗
Figure 8
Figure 8. Figure 8: AIMI v2 precision/recall of DOCTOR 1.00/1.00, ID 1.00/1.00, HOSPITAL 0.94/0.99) view at source ↗
Figure 9
Figure 9. Figure 9: Span-level radar comparison of four transformer models on i2b2 2014 (cross-dataset). view at source ↗
Figure 10
Figure 10. Figure 10: Span-level radar comparison of four transformer models on AIMI (cross-dataset). AIMI view at source ↗

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

40 extracted references · 33 canonical work pages · 3 internal anchors

  1. [1]

    ModernBERT or DeBERTaV3 ? Examining architecture and data influence on transformer encoder models performance

    Wissam Antoun, Beno \^i t Sagot, and Djam \'e Seddah. ModernBERT or DeBERTaV3 ? Examining architecture and data influence on transformer encoder models performance. In Proceedings of the 14th International Joint Conference on Natural Language Processing and the 4th Conference of the Asia-Pacific Chapter of the Association for Computational Linguistics (IJ...

    2025

  2. [2]

    Fr\' e chet distance for offline evaluation of information retrieval systems with sparse labels

    Negar Arabzadeh and Charles Clarke. Fr\' e chet distance for offline evaluation of information retrieval systems with sparse labels. In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers), pages 420--431, St. Julian's, Malta, 2024. Association for Computational Linguistics. UR...

    2024

  3. [3]

    Small Language Models are the Future of Agentic AI

    Peter Belcak, Greg Heinrich, Shizhe Diao, Yonggan Fu, Xin Dong, Saurav Muralidharan, Yingyan Celine Lin, and Pavlo Molchanov. Small Language Models are the Future of Agentic AI , June 2025. URL http://arxiv.org/abs/2506.02153. arXiv:2506.02153 [cs]

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  4. [4]

    The stanford medicine data science ecosystem for clinical and translational research

    Alison Callahan, Euan Ashley, Somalee Datta, Priyamvada Desai, Todd A Ferris, Jason A Fries, Michael Halaas, Curtis P Langlotz, Sean Mackey, José D Posada, Michael A Pfeffer, and Nigam H Shah. The stanford medicine data science ecosystem for clinical and translational research. JAMIA Open, 6 0 (3): 0 ooad054, 08 2023. ISSN 2574-2531. doi:10.1093/jamiaopen...

    work page doi:10.1093/jamiaopen/ooad054 2023

  5. [5]

    Automated deidentification of radiology reports combining transformer and ``hide in plain sight'' rule-based methods

    Pierre J Chambon, Christopher Wu, Jackson M Steinkamp, Jason Adleberg, Tessa S Cook, and Curtis P Langlotz. Automated deidentification of radiology reports combining transformer and ``hide in plain sight'' rule-based methods. Journal of the American Medical Informatics Association, 30 0 (2): 0 318--328, January 2023. ISSN 1067-5027, 1527-974X. doi:10.1093...

    work page doi:10.1093/jamia/ocac219 2023

  6. [6]

    In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)

    Min Jin Chong and David Forsyth. Effectively unbiased FID and inception score and where to find them. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 6070--6079, 2020. doi:10.1109/CVPR42600.2020.00611

    work page doi:10.1109/cvpr42600.2020.00611 2020

  7. [7]

    Exploiting vocabulary frequency imbalance in language model pre-training, 2025

    Woojin Chung and Jeonghoon Kim. Exploiting vocabulary frequency imbalance in language model pre-training, 2025. URL https://arxiv.org/abs/2508.15390. NeurIPS 2025

    work page arXiv 2025

  8. [8]

    A new paradigm for accelerating clinical data science at stanford medicine, 2020

    Somalee Datta, Jose Posada, Garrick Olson, Wencheng Li, Ciaran O'Reilly, Deepa Balraj, Joseph Mesterhazy, Joseph Pallas, Priyamvada Desai, and Nigam Shah. A new paradigm for accelerating clinical data science at stanford medicine, 2020. URL https://arxiv.org/abs/2003.10534

    work page arXiv 2020

  9. [9]

    D. C. Dowson and B. V. Landau. The Fr\' e chet distance between multivariate normal distributions. Journal of Multivariate Analysis, 12 0 (3): 0 450--455, 1982. doi:10.1016/0047-259X(82)90077-X

    work page doi:10.1016/0047-259x(82)90077-x 1982

  10. [10]

    DuVall, and Patrick R

    Hannah Eyre, Qiwei Gan, Mengke Hu, Annie Bowles, Johnathan Stanley, Jianlin Shi, Scott L. DuVall, and Patrick R. Alba. Evaluating Clinical Note Deidentification Tools and Transformer Transferability between Public and Private Data from the US Department of Veterans Affairs , June 2025. URL https://www.medrxiv.org/content/10.1101/2025.03.21.25323520v2

    work page doi:10.1101/2025.03.21.25323520v2 2025

  11. [11]

    Vertex AI — Generative AI Pricing

    Google Cloud . Vertex AI — Generative AI Pricing . https://cloud.google.com/vertex-ai/generative-ai/pricing, 2025. Gemini 2.5 Flash Flex/Batch pricing: \ 0.15 per 1M input tokens, \ 1.25 per 1M output tokens. Accessed March 2025

    2025

  12. [12]

    DeBERTaV3: Improving DeBERTa using ELECTRA-Style Pre-Training with Gradient-Disentangled Embedding Sharing

    Pengcheng He, Jianfeng Gao, and Weizhu Chen. Debertav3: Improving deberta using electra-style pre-training with gradient-disentangled embedding sharing. arXiv preprint arXiv:2111.09543, 2021

    work page internal anchor Pith review Pith/arXiv arXiv 2021

  13. [13]

    An Extensible Evaluation Framework Applied to Clinical Text Deidentification Natural Language Processing Tools : Multisystem and Multicorpus Study

    Paul M Heider and St \'e phane M Meystre. An Extensible Evaluation Framework Applied to Clinical Text Deidentification Natural Language Processing Tools : Multisystem and Multicorpus Study . Journal of Medical Internet Research, 26: 0 e55676, May 2024. ISSN 1439-4456. doi:10.2196/55676. URL https://pmc.ncbi.nlm.nih.gov/articles/PMC11167315/

    work page doi:10.2196/55676 2024

  14. [14]

    GANs trained by a two time-scale update rule converge to a local Nash equilibrium

    Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter. GANs trained by a two time-scale update rule converge to a local Nash equilibrium. In Advances in Neural Information Processing Systems, volume 30, 2017

    2017

  15. [15]

    Comeau, Lana Yeganova, W

    Qiao Jin, Won Kim, Qingyu Chen, Donald C. Comeau, Lana Yeganova, W. John Wilbur, and Zhiyong Lu. MedCPT : Contrastive Pre -trained Transformers with large-scale PubMed search logs for zero-shot biomedical information retrieval. Bioinformatics, 39 0 (11), 2023. doi:10.1093/bioinformatics/btad651

    work page doi:10.1093/bioinformatics/btad651 2023

  16. [16]

    Aman Kansal, Emma Chen, Boyang Tom Jin, Pranav Rajpurkar, and David A. Kim. MC-MED , multimodal clinical monitoring in the emergency department. Scientific Data, 12 0 (1): 0 1094, 2025. doi:10.1038/s41597-025-05419-5. URL https://doi.org/10.1038/s41597-025-05419-5

    work page doi:10.1038/s41597-025-05419-5 2025

  17. [17]

    Wiest, Stephen Gilbert, and Daniel Truhn

    Jakob Nikolas Kather, Dyke Ferber, Isabelle C. Wiest, Stephen Gilbert, and Daniel Truhn. Large language models could make natural language again the universal interface of healthcare. Nature Medicine, 30: 0 2708--2710, 2024. doi:10.1038/s41591-024-03259-5

    work page doi:10.1038/s41591-024-03259-5 2024

  18. [18]

    Beyond Accuracy : Automated De - Identification of Large Real - World Clinical Text Datasets , December 2023

    Veysel Kocaman, Hasham Ul Haq, and David Talby. Beyond Accuracy : Automated De - Identification of Large Real - World Clinical Text Datasets , December 2023. URL http://arxiv.org/abs/2312.08495. arXiv:2312.08495 [cs]

    work page arXiv 2023

  19. [19]

    Automated De - Identification , Consistent Obfuscation , and Regulatory Grade Validation of 2 Billion Patient Notes , September 2025

    Veysel Kocaman, Lindsay Mico, Mustafa Aytug Kaya, Nadaa Taiyab, David Talby, Tae Surh, Yuqing Guo, Vivek Tomer, and Robert Kramer. Automated De - Identification , Consistent Obfuscation , and Regulatory Grade Validation of 2 Billion Patient Notes , September 2025. URL https://www.researchsquare.com/article/rs-6867162/v1

    2025

  20. [20]

    De-identification of clinical free text using natural language processing: A systematic review of current approaches

    Aleksandar Kovačević, Bojana Bašaragin, Nikola Milošević, and Goran Nenadić. De-identification of clinical free text using natural language processing: A systematic review of current approaches. Artificial Intelligence in Medicine, 151: 0 102845, May 2024. ISSN 09333657. doi:10.1016/j.artmed.2024.102845. URL https://linkinghub.elsevier.com/retrieve/pii/S0...

    work page doi:10.1016/j.artmed.2024.102845 2024

  21. [21]

    Shah, Richard Dobson, and James Teo

    Zeljko Kraljevic, Anthony Shek, Joshua Au Yeung, Ewart Jonathan Sheldon, Mohammad Al-Agil, Haris Shuaib, Xi Bai, Kawsar Noor, Anoop D. Shah, Richard Dobson, and James Teo. Validating transformers for redaction of text from electronic health records in real-world healthcare, October 2023. URL http://arxiv.org/abs/2310.04468. arXiv:2310.04468 [cs]

    work page arXiv 2023

  22. [22]

    Rachel Kuo, Andrew A. S. Soltan, Ciaran O'Hanlon, Alan Hasanic, David A. Clifton, Collins Gary, Dominic Furniss, and David W. Eyre. Benchmarking transformer-based models for medical record deidentification: A single centre, multi-specialty evaluation, May 2025. URL https://www.medrxiv.org/content/10.1101/2025.05.05.25326979v1

    work page doi:10.1101/2025.05.05.25326979v1 2025

  23. [23]

    Targeted Error Correction in Knowledge Distillation : Small Language Models Surpass GPT , November 2025

    Hee-Jin Lee, Zhen Guo, Luchao Jin, and Morteza Moazami Goudarzi. Targeted Error Correction in Knowledge Distillation : Small Language Models Surpass GPT , November 2025. URL http://arxiv.org/abs/2511.03005. arXiv:2511.03005 [cs]

    work page arXiv 2025

  24. [24]

    RAD : Towards Trustworthy Retrieval - Augmented Multi -modal Clinical Diagnosis , 2025

    Haolin Li, Tianjie Dai, Zhe Chen, Siyuan Du, Jiangchao Yao, Ya Zhang, and Yanfeng Wang. RAD : Towards Trustworthy Retrieval - Augmented Multi -modal Clinical Diagnosis , 2025. URL https://arxiv.org/abs/2509.19980. NeurIPS 2025

    work page arXiv 2025

  25. [25]

    Divergence measures based on the Shannon entropy

    Jianhua Lin. Divergence measures based on the Shannon entropy. IEEE Transactions on Information Theory, 37 0 (1): 0 145--151, 1991. doi:10.1109/18.61115

    work page doi:10.1109/18.61115 1991

  26. [26]

    arXiv preprint arXiv:2303.11032 , year=

    Zhengliang Liu, Yue Huang, Xiaowei Yu, Lu Zhang, Zihao Wu, Chao Cao, Haixing Dai, Lin Zhao, Yiwei Li, Peng Shu, Fang Zeng, Lichao Sun, Wei Liu, Dinggang Shen, Quanzheng Li, Tianming Liu, Dajiang Zhu, and Xiang Li. DeID - GPT : Zero -shot Medical Text De - Identification by GPT -4, December 2023. URL http://arxiv.org/abs/2303.11032. arXiv:2303.11032 [cs]

    work page arXiv 2023

  27. [27]

    Evaluation metrics for headline generation using deep pre-trained embeddings

    Abdul Moeed, Yang An, Gerhard Hagerer, and Georg Groh. Evaluation metrics for headline generation using deep pre-trained embeddings. In Proceedings of the Twelfth Language Resources and Evaluation Conference, pages 1796--1802, Marseille, France, 2020. European Language Resources Association. URL https://aclanthology.org/2020.lrec-1.222/

    2020

  28. [28]

    Neilson, Moniruzzaman Moni, Marcello Nesca, Alexander Singer, and Jennifer E

    Bekelu Negash, Alan Katz, Christine J. Neilson, Moniruzzaman Moni, Marcello Nesca, Alexander Singer, and Jennifer E. Enns. De-identification of Free Text Data containing Personal Health Information : A Scoping Review of Reviews . International Journal of Population Data Science, 8 0 (1), December 2023. ISSN 2399-4908. doi:10.23889/ijpds.v8i1.2153. URL htt...

    work page doi:10.23889/ijpds.v8i1.2153 2023

  29. [29]

    gpt-oss-120b & gpt-oss-20b Model Card

    OpenAI . GPT-OSS : Open -weight models for reasoning, agentic tasks, and versatile developer use cases. arXiv preprint arXiv:2508.10925, 2025. URL https://huggingface.co/openai/gpt-oss-120b. Mixture-of-Experts architecture; GPT-OSS 120B (117B total, 5.1B active parameters) and GPT-OSS 20B (21B total, 3.6B active parameters). Apache 2.0 license

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  30. [30]

    Osborne, Andrew Trotter, Tobias O'Leary, Chris Coffee, Micah D

    John D. Osborne, Andrew Trotter, Tobias O'Leary, Chris Coffee, Micah D. Cochran, Luis Mansilla-Gonzalez, Akhil Nadimpalli, Alex McAnnally, Abdulateef I. Almudaifer, Jeffrey R. Curtis, Salma M. Aly, and Richard E. Kennedy. A Markov Chain Replacement Strategy for Surrogate Identifiers : Minimizing Re - Identification Risk While Preserving Text Reuse . Elect...

    work page doi:10.3390/electronics14193945 2025

  31. [31]

    Improving the Performance of Radiology Report De -identification with Large - Scale Training and Benchmarking Against Cloud Vendor Methods , November 2025

    Eva Prakash, Maayane Attias, Pierre Chambon, Justin Xu, Steven Truong, Jean-Benoit Delbrouck, Tessa Cook, and Curtis Langlotz. Improving the Performance of Radiology Report De -identification with Large - Scale Training and Benchmarking Against Cloud Vendor Methods , November 2025. URL http://arxiv.org/abs/2511.04079. arXiv:2511.04079 [cs]

    work page arXiv 2025

  32. [32]

    RedactOR : An LLM - Powered Framework for Automatic Clinical Data De - Identification , July 2025

    Praphul Singh, Charlotte Dzialo, Jangwon Kim, Sumana Srivatsa, Irfan Bulu, Sri Gadde, and Krishnaram Kenthapadi. RedactOR : An LLM - Powered Framework for Automatic Clinical Data De - Identification , July 2025. URL http://arxiv.org/abs/2505.18380. arXiv:2505.18380 [cs]

    work page arXiv 2025

  33. [33]

    BioClinical ModernBERT: A state-of-the-art long- context encoder for biomedical and clinical NLP

    Thomas Sounack, Joshua Davis, Brigitte Durieux, Antoine Chaffin, Tom J. Pollard, Eric Lehman, Alistair E. W. Johnson, Matthew McDermott, Tristan Naumann, and Charlotta Lindvall. BioClinical ModernBERT : A State -of-the- Art Long - Context Encoder for Biomedical and Clinical NLP , June 2025. URL http://arxiv.org/abs/2506.10896. arXiv:2506.10896 [cs]

    work page arXiv 2025

  34. [34]

    Annotating longitudinal clinical narratives for de-identification: The 2014 i2b2/uthealth corpus

    Amber Stubbs and "O zlem Uzuner. Annotating longitudinal clinical narratives for de-identification: The 2014 i2b2/uthealth corpus. Journal of Biomedical Informatics, 58: 0 S20--S29, 2015. doi:10.1016/j.jbi.2015.07.020. PMID: 26319540

    work page doi:10.1016/j.jbi.2015.07.020 2014

  35. [35]

    Evaluating the State -of-the- Art in Automatic De -identification

    \"O zlem Uzuner, Yuan Luo, and Peter Szolovits. Evaluating the State -of-the- Art in Automatic De -identification. Journal of the American Medical Informatics Association, 14 0 (5): 0 550--563, September 2007. ISSN 1067-5027. doi:10.1197/jamia.M2444. URL https://doi.org/10.1197/jamia.M2444

    work page doi:10.1197/jamia.m2444 2007

  36. [36]

    Jensen- Shannon divergence message-passing for rich-text graph representation learning

    Zuo Wang and Ye Yuan. Jensen- Shannon divergence message-passing for rich-text graph representation learning. arXiv preprint arXiv:2512.20094, 2025. URL https://arxiv.org/abs/2512.20094

    work page arXiv 2025

  37. [37]

    Wiest, Marie-Elisabeth Le mann, Fabian Wolf, Dyke Ferber, Marko Van Treeck, Jiefu Zhu, Matthias P

    Isabella C. Wiest, Marie-Elisabeth Le mann, Fabian Wolf, Dyke Ferber, Marko Van Treeck, Jiefu Zhu, Matthias P. Ebert, Christoph Benedikt Westphalen, Martin Wermke, and Jakob Nikolas Kather. Deidentifying Medical Documents with Local , Privacy - Preserving Large Language Models : The LLM - Anonymizer . NEJM AI, 2 0 (4): 0 AIdbp2400537, March 2025. doi:10.1...

    work page doi:10.1056/aidbp2400537 2025

  38. [38]

    The design of approximation algorithms

    David P Williamson and David B Shmoys. The design of approximation algorithms. Cambridge university press, 2011

    2011

  39. [39]

    A pragmatic note on evaluating generative models with Fréchet inception distance for retinal image synthesis.arXiv preprint arXiv:2502.17160, 2025

    Yuli Wu, Fucheng Liu, R \"u veyda Yilmaz, Henning Konermann, Peter Walter, and Johannes Stegmaier. A pragmatic note on evaluating generative models with Fr\' e chet inception distance for retinal image synthesis. In Proceedings of Medical Imaging with Deep Learning (MIDL), 2026. URL https://arxiv.org/abs/2502.17160

    work page arXiv 2026

  40. [40]

    Hogan, and Yonghui Wu

    Xi Yang, Tianchen Lyu, Qian Li, Chih-Yin Lee, Jiang Bian, William R. Hogan, and Yonghui Wu. A study of deep learning methods for de-identification of clinical notes in cross-institute settings. BMC Medical Informatics and Decision Making, 19 0 (Suppl 5): 0 232, December 2019. ISSN 1472-6947. doi:10.1186/s12911-019-0935-4. URL https://pmc.ncbi.nlm.nih.gov/...

    work page doi:10.1186/s12911-019-0935-4 2019