AI-Driven Generation of Old English: A Framework for Low-Resource Languages

Cristian L\'opez; Javier Mart\'in Arista; Mat\'ias Nu\~nez; Rodrigo Gabriel Salazar Alva

REVIEW 4 major objections 6 minor 17 references

AI-Driven Generation of Old English: A Framework for Low-Resource Languages

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper claims that an English-centric large language model can be turned into a fluent Old English writer by staging LoRA fine-tuning, backtranslation, and a dual-agent pipeline.

desk verdict A credible LoRA-plus-backtranslation pipeline for Old English whose headline numbers cannot be trusted until the authors prove the test set was held out from all training stages and the expert evaluators were independent. read the letter →

arxiv 2507.20111 v1 pith:OJW2G6EE submitted 2025-07-27 cs.CL cs.AI

Rodrigo Gabriel Salazar Alva , Matías Nuñez , Cristian López , Javier Martín Arista This is my paper · ORCID

classification cs.CLcs.AI

keywords low-resourcelanguagesOldEnglishsyntheticdatagenerationLow-RankAdaptation(LoRA)backtranslationmachinetranslationtextdigitalpreservation

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

This paper argues that the scarcity of Old English text is not a hard barrier to generating new Old English with large language models. The authors take an English-centric base model, adapt it to Old English with efficient fine-tuning on four translation-like tasks, expand the training data by translating Old English back into Modern English, and then pair the two directions into synthetic parallel sentences. A two-agent pipeline then drafts Modern English passages and translates them into Old English. Reported results are large jumps in automatic translation scores—BLEU from roughly 26 to over 65 for English-to-Old English—and expert ratings near 9 out of 10 for inflection, word order, and lexical choice, with semantic coherence weaker at 7.8. The broader promise is a replicable recipe for expanding corpora and supporting the study of other endangered or ancient languages.

What carries the argument

The load-bearing mechanism is the staged adaptation of a single English-centric language model rather than the addition of a separate Old English system. Low-Rank Adaptation (LoRA) keeps pretraining feasible by updating only low-rank weight matrices while the rest of the model stays fixed. Four parallel task templates—Old English text completion, Modern English-to-Old English translation, Old English-to-Modern English backtranslation, and Old English word definition—teach the model Old English grammar and vocabulary through its existing English knowledge. Backtranslation then manufactures synthetic parallel sentence pairs from monolingual Old English, and the dual-agent pipeline (an English fragment generator plus the refined Old English translator) scales this into a corpus-generation loop. The mechanism works by converting the scarcity of bilingual Old English data into an abundance of monolingual data that can be turned into parallel training examples.

What would settle it

Ask the released code and dataset to name every Old English sentence used in training, validation, and backtranslation, then recompute BLEU, CHRF, and METEOR on a test set whose sentences are provably disjoint from that pool; if the scores drop toward the base-model range, the reported gains are measurement artifacts. A stricter test is to have Old English scholars who have not seen the outputs rate them blind and flag anachronisms or grammatical errors without knowing the texts are machine-generated.

Watch

Extended reading notes

Core claim

The central claim is that the low-resource bottleneck for Old English can be broken by reframing synthetic corpus expansion as machine translation. The authors start from a general-purpose English-dominant model and adapt it to Old English in two training phases: first, continual domain-adaptive pretraining over four tasks—completing Old English fragments, translating English into Old English, translating Old English into English, and giving English definitions of Old English words—using Low-Rank Adaptation so only a small set of parameters is updated; second, task specialization in which monolingual Old English is backtranslated into Modern English and the resulting pairs are added to the human-annotated parallel data. A final dual-agent stage separates content generation from translation: one component writes fresh Modern English fragments guided by Old English style samples, and a second translates them into Old English. The paper reports that this recipe lifts English-to-Old English BLEU from 25.94 to 65.41, with CHRF 57.82 and METEOR 57.40, and expert ratings of 9.0 for inflection, 9.0 for word order, 9.1 for lexical choice, and 7.8 for semantic coherence.

Load-bearing premise

The framework's reported quality numbers assume that the Old English sentences used for evaluation were never used in training, domain adaptation, or backtranslation, and that the expert ratings reflect independent Old English specialists rather than the authors themselves.

Editorial extensions

If this is right

If the reported scores hold, English-to-Old English translation becomes a practical capability of an English-centric model: BLEU more than doubles relative to the base model, from about 26 to 65.
The backtranslation step buys roughly 6 additional points on top of domain adaptation, so synthetic parallel data is what turns a stagnating adapted model into a fluent translator.
The dual-agent pipeline can generate large volumes of stylistically anchored Old English, meaning corpus expansion for Old English can proceed at scale and only high-rated outputs need be retained.
High expert scores for inflection, word order, and lexical choice (9.0, 9.0, 9.1) with a lower semantic coherence score (7.8) imply that the generated texts are grammatically reliable, but narration-level coherence is the remaining bottleneck.
The same LoRA-plus-backtranslation-plus-dual-agent recipe is proposed as a transferable template for other low-resource or endangered languages with a small surviving corpus and a related modern language.

Reading between the lines

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

Editorial inference: if this framework is adopted for corpus building, synthetic Old English should be labeled as synthetic; otherwise future models trained on the expanded corpus will inherit the translator's own errors and anachronisms, and the corpus will become circular.
Editorial inference: because backtranslation creates training pairs from the model's own reverse translations, the quality ceiling of the synthetic data is set by the reverse-direction model; a natural stress test is to compare outputs trained on backtranslated data from the same model versus a stronger or ensemble reverse translator.
Editorial inference: the paper's semantic-coherence gap (7.8) suggests a concrete next experiment the authors did not run—retrieval-augmented generation grounded in authentic Old English passages should be tested against the current pipeline on coherence ratings, since the authors name RAG as future work but do not evaluate it.
Editorial inference: if the evaluation set is clean and the expert ratings are independent, this recipe may transfer to other ancient languages with related modern descendants, but the quality numbers themselves would need to be re-established for each new language through its own expert evaluation.

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.

Desk Editor's Note

read the letter

You should know two things about arXiv:2507.20111. First, it is a real engineering effort: the authors fine-tune Llama for English–Old English translation using domain-adaptive pretraining, LoRA, and backtranslation, and they report large jumps (BLEU from ~26 to 65 for EN→ANG) plus expert ratings around 9/10 for inflection, word order, and lexical choice. Second, the evaluation hygiene is not yet established, and that is the load-bearing issue.

What is genuinely useful: the staged pipeline itself, with four well-chosen DAPT tasks, synthetic parallel pairs from backtranslation, and a dual-agent generation scheme (GPT-4o-mini for English prompts, then their own translator into Old English). The qualitative examples, especially the error table and the sample outputs in Table 6, show the model's progression and its remaining weaknesses. The limitations section is unusually candid: semantic coherence, anachronisms, V2 avoidance, and co-reference drift are all acknowledged. They also cite the right prior work and provide a GitHub link, though no code or data appears in the manuscript itself.

The soft spots are real but not necessarily fatal. The most serious is the lack of any demonstrated separation between the test set and the data used for domain adaptation, backtranslation, or the synthetic fine-tuning corpus. The paper says the data was split into train/validation/test, but it never shows document-level or n-gram deduplication, and the Phase 2 backtranslation uses "previously unseen fragments from the monolingual Old English corpus"—which may still overlap with test documents. If that overlap exists, the BLEU jump from 25.94 to 65.41 could be memorization, not generalization. Second, the numbers are internally inconsistent: Table 2 gives Llama3.1 8B EN→ANG BLEU 29.13, while Table 5 gives 25.94 for the same direction and model, and the base model is named both Llama-8b and Llama3.1 8B. Third, the expert evaluators are not disclosed, and the PARCOROE benchmark [3] is edited by co-author Martín Arista, so the human scores are not demonstrably independent. Fourth, there are no error bars, rater counts, test-set sizes, or hyperparameter values (LoRA rank, alpha, learning rate, etc.), which makes replication harder than it should be.

These are fixable in revision. The central approach is credible and consistent with prior low-resource MT work, and the authors are not trying to hide the model's shortcomings. But the paper as written overclaims because the evidence for a clean evaluation is missing.

Who should read this: people working on low-resource NLP, ancient language preservation, or digital humanities workflows. A serious referee should engage with it and ask for the missing details—especially dataset split provenance, evaluator independence, and corrected tables—rather than letting it slide. I would accept it for peer review, but I would not cite it as a demonstrated result until those questions are answered.

Referee Report

4 major / 6 minor

Summary. The paper presents a framework for generating Old English text using a multi-stage pipeline: domain-adaptive continual pretraining with LoRA, backtranslation-based data augmentation, and a dual-agent architecture (FragmentGen generating English content and OldEnglishTranslator translating to Old English). The authors report large automated-metric gains (EN→ANG BLEU rising from ~26 to 65.41 in Table 5) and high expert ratings (e.g., 9.0 for inflection, 9.1 for lexical choice in Table 7). The central claim is that the resulting OldEnglishRefined model produces fluent, grammatically accurate Old English suitable for corpus expansion, and that the framework is a replicable blueprint for other low-resource languages.

Significance. If the reported results are valid, the framework is a valuable practical contribution to low-resource language processing and digital humanities. The paper's strengths include a reproducible code repository, a clearly structured pipeline (LoRA, backtranslation, dual-agent prompting), a wide range of qualitative examples, and an explicit discussion of limitations. The reported gains are substantial and the human-evaluation scores are high. However, the significance is currently capped by unresolved evaluation threats: possible test-set leakage across training stages, inconsistent baseline scores across tables, and a human evaluation whose independence is not established. These issues must be resolved before the headline numbers can be trusted.

major comments (4)

[Data Preparation; Phase 2: Task Specialization] The paper does not establish that the evaluation test set is disjoint from all data used in training. In 'Data Preparation', the DOEC and PARCOROE data are split into training, validation, and testing subsets, but no document-level or n-gram-level deduplication is reported. In 'Phase 2: Task Specialization', the model is fine-tuned on synthetic parallel pairs generated via backtranslation from 'previously unseen fragments from the monolingual Old English corpus', and in 'Synthetic Data Generation', the FragmentGen agent receives randomly sampled fragments from the DOEC as contextual anchors. If any of these fragments overlap with test sentences, the jump from BLEU 25.94 to 65.41 (Table 5) could reflect memorization rather than generalization. Please report explicitly whether test sentences were excluded from the monolingual backtranslation source, the few-shot anchors, and the domain-adaptation corpus, and provide overlap statistics (e.g., maximum n-gram overlap, document-level deduplication).
[Tables 2 and 5] The baseline numbers are internally inconsistent. Table 2 reports Llama3.1 8B EN→ANG scores of BLEU 29.13, CHRF 25.51, METEOR 18.95, while Table 5 reports the same baseline (listed as 'Llama') with BLEU 25.94, CHRF 22.97, METEOR 17.43. Additionally, Table 4 shows OldEnglishBase at 3 epochs achieving EN→ANG BLEU 60.73, whereas Table 5 lists OldEnglishBase at BLEU 59.99. These discrepancies make the exact improvement magnitudes ambiguous and should be reconciled, with a clear statement of which configuration (checkpoint, epoch, evaluation set) each table reports.
[Expert Linguistic Evaluation; Table 7] The human evaluation lacks the transparency needed to support the reported high scores. The text mentions 'expert reviewers' and 'specialist linguists', but does not report the number of raters, their affiliations, whether they were independent of the authors, inter-rater agreement, or the number of generated texts rated per criterion. This matters particularly because the PARCOROE corpus [3] used for evaluation is edited by co-author J. Martín Arista, so the evaluation is not demonstrably independent. Please provide full details of the evaluation protocol and, ideally, an external evaluation by linguists not involved in the project.
[Results and Discussion; Tables 4 and 5] Automated scores are reported without error bars, significance tests, or characteristics of the test set (e.g., number of sentences, source corpus, genre distribution). Given the 2–4 point differences that are used to support claims about backtranslation gains and model specialization, the absence of variance estimates makes it difficult to assess whether these differences are meaningful. Please add test-set statistics and, if possible, bootstrap confidence intervals or multiple evaluation runs.

minor comments (6)

[Throughout] The base model is referred to inconsistently as 'Llama-8b' and 'Llama3.1 8B'; please use a single, precise designation (with the parameter count, e.g., Llama-3.1-8B).
[Table 2] The header contains 'Modelo' (Spanish) instead of 'Model'.
[Table 6] The source sentence contains 'except.the’angild’to others, aud fortress-work', where 'aud' appears to be a typo for 'and'.
[Figure 3] The figure caption uses 'OE-Base' and 'OE-Refined', while the text uses 'OldEnglishBase' and 'OldEnglishRefined'; please align the terminology.
[Abstract] The phrase 'BLEU scores increasing from 26 to over 65' is consistent with Table 5 but not with Table 2's baseline of 29.13; clarify which baseline is referenced.
[Results and Discussion] There are minor typos such as 'demostrated' (should be 'demonstrated') and inconsistent capitalization of 'Contemporary English' / 'Modern English'.

Circularity Check

1 steps flagged · score 4.0 of 10

The training pipeline is standard fine-tuning, but the evaluation chain is partly self-referential: the PARCOROE resource used for training and reference translations is edited by co-author Martín Arista, and the expert raters' independence is never disclosed.

self citation load bearing [Data Preparation; Reference [3]; Expert Linguistic Evaluation (Table 7)]
"Additionally, an annotated subcorpus of the DOEC was employed to provide the model with examples of translations [3]."

The model is trained on the PARCOROE parallel corpus, and the paper's central evidence of quality is expert ratings ('expert linguists' / 'specialist linguists familiar with Old English') plus references used by BLEU/CHRF/METEOR. Reference [3] is Parcoroev3, edited by co-author J. Martín Arista. The paper never discloses who performed the expert evaluation, whether the raters were independent of PARCOROE, or whether the held-out references came from a source external to the training resource. The headline claim that expert assessment confirms high grammatical and lexical quality is therefore not shown to be anchored outside the authors' own resource/group; the evaluation is partly self-referential rather than external confirmation.

full rationale

This is an empirical ML paper, not a mathematical derivation, so the strict Eq.-to-Eq. circularity test applies mainly to the evaluation setup. The training pipeline (LoRA DAPT, backtranslation, dual-agent generation) is standard and does not reduce to its inputs by definition: the reported BLEU/CHRF/METEOR gains are computed against human reference translations. However, the evaluation credibility is weakened by a load-bearing self-reference: the annotated parallel corpus [3] used to provide training examples is edited by co-author J. Martín Arista, and the expert human evaluators are never identified or shown to be independent of that same project. The authors' own limitation section concedes that 'further validation by expert reviewers and systematic comparison with extant Old English corpora are essential,' which undercuts the strength of the Table 7 expert scores. Potential test-set leakage from backtranslation (monolingual DOEC fragments used for synthetic parallel data and evaluation references from the same corpus) is an additional validation risk, but the paper does not provide enough detail to prove overlap, so I do not count it as demonstrated circularity. Overall: one load-bearing self-referential evaluation chain; the central claim still has independent automated-metric content, so score 4, not 6-10.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The framework rests on corpus quality, metric validity, the fixed behavior of a proprietary model, and the exact identity of the base model; these are assumed rather than demonstrated.

free parameters (1)

Fine-tuning hyperparameters (LoRA rank, alpha, learning rate, batch size, epochs) = not reported
The paper describes LoRA and DAPT but gives no numerical configuration; the reported performance depends on these unstated choices, so they are free parameters with unspecified values.

assumptions (4)

domain assumption The DOEC and Bosworth-Toller dictionary are accurate and representative sources of Old English.
Invoked in Data Preparation; if the corpus is small or non-representative, the model and evaluation inherit that bias.
domain assumption BLEU, METEOR, and CHRF scores correlate with human-judged quality for Old English machine translation.
Invoked in Evaluation; these metrics are normally tuned on modern languages, and their validity on a synthetic low-resource historical language is assumed.
domain assumption GPT-4o-mini can reliably produce stylistically appropriate Modern English fragments from Old English examples.
The FragmentGen agent is a proprietary model whose behavior is fixed and not auditable; this is assumed in Synthetic Data Generation.
domain assumption The base model is an 8B-parameter Llama model with English-centric capabilities as claimed.
The paper cites the Llama 3 herd paper but names the checkpoint 'Llama-8b' without a precise version, yet that model is the foundation of the pipeline.

how reviews work

0 comments

Cite this review

Pith. "Pith review of AI-Driven Generation of Old English: A Framework for Low-Resource Languages." pith.science (2026). https://pith.science/paper/OJW2G6EE

@misc{pith2026250720111,
  author       = {Pith},
  title        = {Pith review of: AI-Driven Generation of Old English: A Framework for Low-Resource Languages},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OJW2G6EE}},
  note         = {Machine review of arXiv:2507.20111}
}

read the original abstract

Preserving ancient languages is essential for understanding humanity's cultural and linguistic heritage, yet Old English remains critically under-resourced, limiting its accessibility to modern natural language processing (NLP) techniques. We present a scalable framework that uses advanced large language models (LLMs) to generate high-quality Old English texts, addressing this gap. Our approach combines parameter-efficient fine-tuning (Low-Rank Adaptation, LoRA), data augmentation via backtranslation, and a dual-agent pipeline that separates the tasks of content generation (in English) and translation (into Old English). Evaluation with automated metrics (BLEU, METEOR, and CHRF) shows significant improvements over baseline models, with BLEU scores increasing from 26 to over 65 for English-to-Old English translation. Expert human assessment also confirms high grammatical accuracy and stylistic fidelity in the generated texts. Beyond expanding the Old English corpus, our method offers a practical blueprint for revitalizing other endangered languages, effectively uniting AI innovation with the goals of cultural preservation.

Figures

Figures reproduced from arXiv: 2507.20111 by the authors.

**Figure 2.** FIGURE 2 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗

**Figure 3.** FIGURE 3 [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

17 extracted references · 7 canonical work pages

[3]

Parcoroev3: An open access annotated parallel corpus old english- english,

J. M. Arista, S. D. Barrag ´an, L. F. Allo, L. G. Fern ´andez, Y . H. Bghiyel, M. L. Palacios, R. M. Mendaza, C. N. Urraca, A. E. O. L´opez, E. L. R. Narbona, R. T. Alonso, and R. V . Escarza, “Parcoroev3: An open access annotated parallel corpus old english- english,” Nerthus Project, Universidad de La Rioja, 2023, edited by J. Mart´ın Arista. [Online]. ...

work page 2023
[1]

Machine learning for ancient languages: A survey,

T. Sommerschield, Y . Assael, J. Pavlopoulos, V . Stefanak, A. Senior, C. Dyer, J. Bodel, J. Prag, I. Androutsopoulos, and N. de Freitas, “Machine learning for ancient languages: A survey,” Computational Linguistics, vol. 49, no. 3, pp. 703–747, 2023

work page 2023
[2]

The dictionary of old english web corpus,

J. P. W. A. diPaolo Healey and X. Xiang., “The dictionary of old english web corpus,” http://www.doe.utoronto.ca/, 2004

work page 2004
[4]

Bosworth and T

J. Bosworth and T. N. Toller, An Anglo-Saxon Dictionary . Oxford: Clarendon Press, 1972, supplement by T. N. Toller; Addenda and corrigenda by A. Campbell; Digital version by O. Tichy and M. Roceck. [Online]. Available: https://bosworthtoller.com

work page 1972
[5]

The Classical Language Toolkit: An NLP framework for pre-modern languages,

K. P. Johnson, P. J. Burns, J. Stewart, T. Cook, C. Besnier, and W. J. B. Mattingly, “The Classical Language Toolkit: An NLP framework for pre-modern languages,” in Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing: System Demonstrations, H. Ji...

work page 2021
[6]

Don’t stop pretraining: Adapt language models to domains and tasks,

S. Gururangan, A. Marasovi ´c, S. Swayamdipta, K. Lo, I. Beltagy, D. Downey, and N. A. Smith, “Don’t stop pretraining: Adapt language models to domains and tasks,” 2020. [Online]. Available: https://arxiv.org/abs/2004.10964

arXiv 2020
[7]

Efficient continual pre-training for building domain specific large language models,

Y . Xie, K. Aggarwal, and A. Ahmad, “Efficient continual pre-training for building domain specific large language models,” in Findings of the Association for Computational Linguistics: ACL 2024 , L.-W. Ku, A. Martins, and V . Srikumar, Eds. Bangkok, Thailand: Association for Computational Linguistics, Aug. 2024, pp. 10 184–10 201. [Online]. Available: htt...

work page 2024
[8]

The llama 3 herd of models,

A. Dubey et al. , “The llama 3 herd of models,” 2024. [Online]. Available: https://arxiv.org/abs/2407.21783

arXiv 2024

Show all 17 references

[9]

Lora: Low-rank adaptation of large language models,

E. J. Hu, Y . Shen, P. Wallis, Z. Allen-Zhu, Y . Li, S. Wang, L. Wang, and W. Chen, “Lora: Low-rank adaptation of large language models,” 2021

2021
[10]

Understanding back-translation at scale,

S. Edunov, M. Ott, M. Auli, and D. Grangier, “Understanding back-translation at scale,” 2018. [Online]. Available: https://arxiv.org/ abs/1808.09381

2018 arXiv
[11]

Democratizing llms for low-resource languages by leveraging their english dominant abilities with linguistically-diverse prompts,

X.-P. Nguyen, S. M. Aljunied, S. Joty, and L. Bing, “Democratizing llms for low-resource languages by leveraging their english dominant abilities with linguistically-diverse prompts,” 2023. [Online]. Available: https://arxiv.org/abs/2306.11372v1

2023 arXiv
[12]

Bleu: a method for automatic evaluation of machine translation,

K. Papineni, S. Roukos, T. Ward, and W.-J. Zhu, “Bleu: a method for automatic evaluation of machine translation,” in Proceedings of the 40th annual meeting of the Association for Computational Linguistics , 2002, pp. 311–318

2002
[13]

Meteor: An automatic metric for mt evaluation with improved correlation with human judgments,

S. Banerjee and A. Lavie, “Meteor: An automatic metric for mt evaluation with improved correlation with human judgments,” in Proceedings of the acl workshop on intrinsic and extrinsic evaluation measures for machine translation and/or summarization , 2005, pp. 65–72

2005
[14]

chrf: character n-gram f-score for automatic mt evalu- ation,

M. Popovi ´c, “chrf: character n-gram f-score for automatic mt evalu- ation,” in Proceedings of the tenth workshop on statistical machine translation, 2015, pp. 392–395

2015
[15]

Mistral 7b,

A. Q. Jiang, A. Sablayrolles, A. Mensch, C. Bamford, D. S. Chaplot, D. de las Casas, F. Bressand, G. Lengyel, G. Lample, L. Saulnier, L. R. Lavaud, M.-A. Lachaux, P. Stock, T. L. Scao, T. Lavril, T. Wang, T. Lacroix, and W. E. Sayed, “Mistral 7b,” 2023. [Online]. Available: ht...

2023 arXiv
[16]

Retrieval-augmented generation for knowledge-intensive nlp tasks,

P. Lewis, E. Perez, A. Piktus, F. Petroni, V . Karpukhin, N. Goyal, H. K ¨uttler, M. Lewis, W. tau Yih, T. Rockt ¨aschel, S. Riedel, and D. Kiela, “Retrieval-augmented generation for knowledge-intensive nlp tasks,” in Advances in Neural Information Processing Systems (NeurIPS)...

2020
[17]

Improving retrieval-augmented large language models via data importance learning,

X. Lyu, S. Grafberger, S. Biegel, S. Wei, M. Cao, S. Schelter, and C. Zhang, “Improving retrieval-augmented large language models via data importance learning,” arXiv preprint arXiv:2307.03027 , 2023. [Online]. Available: https://arxiv.org/abs/2307.03027 10

2023 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [3]

Parcoroev3: An open access annotated parallel corpus old english- english,

J. M. Arista, S. D. Barrag ´an, L. F. Allo, L. G. Fern ´andez, Y . H. Bghiyel, M. L. Palacios, R. M. Mendaza, C. N. Urraca, A. E. O. L´opez, E. L. R. Narbona, R. T. Alonso, and R. V . Escarza, “Parcoroev3: An open access annotated parallel corpus old english- english,” Nerthus Project, Universidad de La Rioja, 2023, edited by J. Mart´ın Arista. [Online]. ...

work page 2023

[2] [1]

Machine learning for ancient languages: A survey,

T. Sommerschield, Y . Assael, J. Pavlopoulos, V . Stefanak, A. Senior, C. Dyer, J. Bodel, J. Prag, I. Androutsopoulos, and N. de Freitas, “Machine learning for ancient languages: A survey,” Computational Linguistics, vol. 49, no. 3, pp. 703–747, 2023

work page 2023

[3] [2]

The dictionary of old english web corpus,

J. P. W. A. diPaolo Healey and X. Xiang., “The dictionary of old english web corpus,” http://www.doe.utoronto.ca/, 2004

work page 2004

[4] [4]

Bosworth and T

J. Bosworth and T. N. Toller, An Anglo-Saxon Dictionary . Oxford: Clarendon Press, 1972, supplement by T. N. Toller; Addenda and corrigenda by A. Campbell; Digital version by O. Tichy and M. Roceck. [Online]. Available: https://bosworthtoller.com

work page 1972

[5] [5]

The Classical Language Toolkit: An NLP framework for pre-modern languages,

K. P. Johnson, P. J. Burns, J. Stewart, T. Cook, C. Besnier, and W. J. B. Mattingly, “The Classical Language Toolkit: An NLP framework for pre-modern languages,” in Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing: System Demonstrations, H. Ji...

work page 2021

[6] [6]

Don’t stop pretraining: Adapt language models to domains and tasks,

S. Gururangan, A. Marasovi ´c, S. Swayamdipta, K. Lo, I. Beltagy, D. Downey, and N. A. Smith, “Don’t stop pretraining: Adapt language models to domains and tasks,” 2020. [Online]. Available: https://arxiv.org/abs/2004.10964

arXiv 2020

[7] [7]

Efficient continual pre-training for building domain specific large language models,

Y . Xie, K. Aggarwal, and A. Ahmad, “Efficient continual pre-training for building domain specific large language models,” in Findings of the Association for Computational Linguistics: ACL 2024 , L.-W. Ku, A. Martins, and V . Srikumar, Eds. Bangkok, Thailand: Association for Computational Linguistics, Aug. 2024, pp. 10 184–10 201. [Online]. Available: htt...

work page 2024

[8] [8]

The llama 3 herd of models,

A. Dubey et al. , “The llama 3 herd of models,” 2024. [Online]. Available: https://arxiv.org/abs/2407.21783

arXiv 2024