Pith. sign in

REVIEW 4 major objections 6 minor 25 references

GovRelBench:A Benchmark for Government Domain Relevance

T0 review · 4 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read The paper claims that government-domain relevance is a continuous spectrum that a small encoder can score precisely, making it a reusable evaluation tool for LLMs.

desk verdict A useful niche benchmark whose central relevance scores are the authors' own subjective mapping, honestly acknowledged in the paper's limitations. read the letter →

arxiv 2507.21419 v1 pith:EYVDVWPL submitted 2025-07-29 cs.AI

Haiquan Wang , Yi Chen , Shang Zeng , Yun Bian , Zhe Cui This is my paper
classification cs.AI
keywords governmentdomainrelevanceLLMevaluationbenchmarksoftlabelsBetadistributiondiffusionModernBERTscoringChineseregression
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

The paper aims to make government-domain relevance a measurable, comparable quantity for large language models rather than a vague impression. It proposes GovRelBench, a set of 50 prompts that probe how well an LLM stays on government topics, plus GovRelBERT, a 149M-parameter encoder trained to score any text on a 0-to-1 government-relevance scale. The training recipe, SoftGovScore, turns coarse category labels into hand-set scores such as Government=1.00, News=0.70, and Others=0.00, then diffuses each score through a Beta distribution so the model learns graded relevance instead of crisp classes. If those scores are valid, GovRelBench gives researchers a fast, reproducible way to rank and compare LLMs for government-domain work, filling a gap left by safety-focused evaluations.

What carries the argument

The machinery is SoftGovScore's two-step label transform, pinned by Equation (4): each category label L is mapped to a hard score H_L in [0,1], then soft scores are sampled from a Beta distribution whose mean equals H_L because alpha/(alpha+beta)=H_L, with alpha+beta=20 to keep prior confidence uniform while letting variance shrink near 0 and 1 and grow near 0.5. GovRelBERT is a ModernBERT encoder (long bidirectional context, 8192-token window, 149M parameters) with a single regression head trained by mean squared error on those soft targets. The Beta diffusion is the load-bearing step: it turns discrete hand-set reference points into a smooth, uncertainty-aware training distribution.

What would settle it

Give a fresh panel of domain experts, who had no role in Table 4, a new set of 50 government-domain prompts, have them rank LLM outputs by relevance, and check whether GovRelBERT's continuous scores order the outputs the same way; a near-zero correlation, or experts judging highly scored texts as off-topic, would refute the claim.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that direct regression onto soft relevance scores works where classification fails: binary classifiers and embedding-based baselines stay near 24-50% accuracy at separating government from non-government text, while GovRelBERT reaches 71.5% accuracy within a 0.1 error tolerance and 93.6% within 0.2 on the held-out test set. The same scorer then ranks LLMs: Chinese-focused models such as deepseek-chat (0.7959 mean relevance) and Qwen2.5-72B-Instruct (0.7967) score well above Claude-3-7-sonnet (0.4533), with GPT-4o in between (0.7359). The authors conclude that domain relevance is best treated as a continuous spectrum and that a small encoder trained with SoftGovScore can serve as the evaluation tool for a government-domain benchmark.

Load-bearing premise

The load-bearing premise is that the hand-assigned hard scores in Table 4 (Government=1.00, Law_A=0.90, News=0.70, Soy Milk=0.10, Others=0.00) together with Beta diffusion at alpha+beta=20 are a valid ground truth for government-domain relevance; if they do not reflect expert judgment on free-form LLM outputs, every score in Tables 1 and 3 inherits the error.

Editorial extensions

If this is right

  • GovRelBench supplies a reusable 50-prompt test: any LLM can be scored on government-domain relevance by feeding its responses to GovRelBERT, with no additional annotation required.
  • GovRelBERT's small size and speed make the benchmark practical to run repeatedly, for example during model selection or fine-tuning, on a single consumer GPU.
  • The reported rankings show the benchmark can separate model families: Qwen2.5-72B-Instruct and deepseek-chat lead with means near 0.80, while Claude-3-7-sonnet trails at 0.45.
  • Direct score regression (GovRelBERT A) is the recommended evaluation configuration, beating the classify-then-score variant in the 0.1-tolerance comparison and the human panel's 3:2 preference.

Reading between the lines

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

  • SoftGovScore should transfer to other fuzzy domains such as law, finance, or medicine, since the recipe only assumes that anchor scores can be set; the paper has validated it only for government, so this is an extrapolation.
  • The variance column in Table 3 can double as a consistency diagnostic: a model with low variance stays on-topic across all 50 prompts, while high variance flags prompts where the model drifts, a use the paper does not develop.
  • A natural stress test would be to apply GovRelBERT to outputs from prompts outside the 50-set and compare against expert relevance ratings; strong agreement would show the label premise generalizes beyond curated test prompts.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The paper proposes GovRelBench, a benchmark for evaluating LLMs in the Chinese government domain, together with GovRelBERT, a ModernBERT-based regression model that assigns a continuous "government domain relevance" score to text. The key methodological contribution is SoftGovScore, a two-step procedure that first maps 18 domain categories to hand-assigned hard relevance scores (e.g., Government=1.0, News=0.7, Soy Milk=0.1) and then diffuses these hard scores into soft training targets by sampling from Beta distributions whose mean equals the hard score (Eq. 4). GovRelBERT is trained with MSE loss on these soft targets. The paper reports that GovRelBERT substantially outperforms traditional classifiers and larger LLMs on the in-distribution test set (Table 1), and applies the model to score responses of several LLMs to 50 government-domain prompts (Table 3). The central claim is that GovRelBERT accurately computes government-domain relevance and that GovRelBench provides a quantitative metric for LLM evaluation in this domain.

Significance. If the relevance scores were externally validated, the paper would offer a useful and inexpensive evaluation tool for a relatively underserved domain: a reproducible 50-prompt test set, a fast 149M-parameter scorer, a clear training pipeline, and public code/data. The authors also provide an ablation (GovRelBERT B) and report the random seed, which are praiseworthy. However, the significance is currently limited by the fact that the entire evaluation loop is anchored to the authors' own subjective hard-score mapping, and none of the headline numbers are checked against independent human judgments of relevance. The contribution is therefore more a demonstration of a plausible training procedure than a validated benchmark.

major comments (4)
  1. [§4.2 / Appendix A.2 / Eq. (4)] The entire training-target distribution is anchored to the authors' hand-assigned category scores in Table 4, with small boundary adjustments described in §4.2, through Eq. (4), which fixes the Beta mean to the hard score. There is no inter-annotator agreement, no external expert elicitation protocol, and no validation against human judgments of actual LLM responses. Because Table 1's accuracy/F1 numbers are computed against these soft targets and Table 3's model rankings are produced by the resulting scorer, any error in this mapping propagates to every headline result. The paper's own Section 5 concedes that the mapping relies heavily on subjective judgment. This is load-bearing for the abstract's claim that GovRelBERT "accurately computes" relevance and for §3.3's claim that GovRelBench provides a "quantitative metric." Please add external human relevance ratings on a sample of training texts and on the actual LLM responses, report agreement/correlation with GovRelBERT scores, and ideally include a sensitivity analysis over plausible alternative mappings.
  2. [§4.3.2 / Table 1] The evaluation of GovRelBERT is performed on a held-out subset of the same dataset whose soft labels were generated by the authors' Beta diffusion from their own hard scores. Thus the high accuracy/F1 values (e.g., 93.58% at tolerance 0.2) primarily show that GovRelBERT can reproduce the authors' labeling convention on in-distribution documents; they do not establish accuracy on arbitrary government-domain text, and they certainly do not validate the relevance scores assigned to free-form LLM responses. In addition, all values in Table 1 are single-run point estimates with no confidence intervals, repeated seeds, or significance tests, which makes the claimed superiority over much larger LLMs difficult to assess. Please provide multiple seeds/error bars and, more importantly, an independently annotated evaluation set with human relevance scores.
  3. [§4.4 / Table 3 / §4.3.3] The benchmark application in §4.4 scores free-form LLM responses with GovRelBERT, but no human rating of those responses is reported. The human panel in §4.3.3 compared GovRelBERT A and B on selected high- and low-error samples; it did not check whether the target scores themselves are correct. The LLM-judge panel in Table 2 is also weak evidence: five judges split 2-2-1, the per-instance vote count is 11 vs 9, and no judge prompts, sampling details, or agreement statistics are provided. Consequently, Table 3's model rankings are not grounded in an independent measure of government-domain relevance. Please add human judgments on the actual 50-prompt responses, with inter-annotator agreement and correlation between GovRelBERT scores and human scores, and report at least a few qualitative examples of scores assigned to generated responses.
  4. [§3.3] GovRelBench currently consists of only 50 self-authored prompts (45 authored plus 5 news-like), with no stated selection criteria, no analysis of coverage across government tasks, and no stability check. For a reusable benchmark, the authors should describe how prompts were constructed and selected, characterize the covered government scenarios, and test whether model rankings are stable under prompt rewording or repeated sampling. Without this, the benchmark's reliability as a quantitative evaluation instrument is unestablished.
minor comments (6)
  1. [§3.1 / Eq. (4)] The notation E[beta_L(α, β)] in Eq. (4) is confusing because β is both the distribution parameter and part of the distribution name; please use E[S_soft | α, β] or similar.
  2. [§4.2] The boundary adjustments ('Government' to approximately 0.95 and 'Other' to approximately 0.15) are described only in prose; please provide the final hard-score mapping table so that the training targets are fully specified.
  3. [Table 1] Please define how Accuracy and F1 are computed for a continuous scorer with tolerance thresholds, and clarify that 'GovRelBERT A (Error Tolerance 0.1)' is an evaluation setting rather than a separate model.
  4. [§4.4 / Table 3] There are naming inconsistencies: the text mentions Qwen1.5-72B-Instruct and Claude-3 Opus while Table 3 lists Qwen2.5-72B-Instruct and claude-3-7-sonnet; please align names and versions. The speculative sentence about a hypothetical 115B TeleChat variant should be removed or clearly labeled as speculation.
  5. [§4.3.1] The optimal model being reached at approximately 0.7 epochs is unusual; please explain the early-stopping configuration, the patience setting, and why the optimum occurs before one full epoch.
  6. [§4.3.3] The text contains formatting artifacts such as 'M odernBERTA' and 'M odernBERTB'; please fix these. Also provide the exact prompts and protocol used for the LLM judges.

Circularity Check

2 steps flagged · score 6.0 of 10

GovRelBERT's relevance scores inherit their ground truth from the authors' own subjective category-to-score mapping: Eq. (4) fixes every training target's mean to the hand-set H_L, so Table 1's accuracy gauges fit to that convention and Table 3's LLM rankings are predictions of the fitted labels.

  1. self definitional [Section 3.1, Eq. (1) and Eq. (4); Section 4.2; Appendix A.2, Table 4; Abstract]
    "By utilizing the formula for the expectation of the Beta distribution, we determine the alpha and beta parameters for each class such that the following equation is satisfied: E[betaL(α, β)] = α/(α+β) = HL (4) ... This involved setting anchor points for clear cases(e.g., Government ≈ 1, News ≈ 0.7, unrelated ≈ 0) and subjectively assigning scores to intermediate categories."

    The only definition of 'government domain relevance score' is the authors' own hard-score mapping in Appendix A.2, Table 4 (Government 1.00, Law_A 0.90, News 0.70, Soy Milk 0.10, Others 0.00). Section 4.2 concedes these scores were subjectively assigned. Eq. (1) sets H_L = f_map(L); Eq. (4) forces the Beta target mean to equal H_L; Eq. (5) trains GovRelBERT by MSE on samples from that Beta; Section 4.3.2 computes Table 1's 71.47%/93.58% accuracy/F1 against targets generated by this same pipeline. The Abstract's claim that GovRelBERT 'accurately compute[s] the text's government domain relevance score' is thus validated only against a ground truth that is itself the training input: the predicted score is defined by the very mapping the model is trained to reproduce.

  2. fitted input called prediction [Section 4.4, evaluation application; Section 3.3; Section 4.3.2; Section 5 (Limitations)]
    "Our trained GovRelBERT model (GovRelBERT A) was then used to assess each response, assigning it a governmental relevance score based on the methodology described earlier. This process allows for a quantitative comparison of how well different LLMs generate content relevant to the governmental domain."

    Section 4.4 feeds free-form LLM responses on the 50 GovRelBench prompts to GovRelBERT and presents the outputs as a measure of 'how well different LLMs generate content relevant to the governmental domain'; Section 3.3 calls this 'a quantitative metric for evaluating an LLM's performance in the government domain.' But GovRelBERT was trained on soft-score targets whose means are the subjectively fitted hard scores (Eq. 4). Every Table 3 score and model ranking is therefore a prediction of that fitted convention, extrapolated to new texts, not a measurement anchored to an independent gold standard; the model cannot output anything independent of the fitted labels.

full rationale

Two circular steps close the evaluation loop. Step 1 (self-definitional): the target quantity 'government domain relevance score' is defined by the authors' own subjective category-to-score mapping (Table 4), and Eq. (4) pins the Beta-diffusion training targets' mean to that mapping, so Table 1's 'accurate' prediction figures measure in-distribution fit to the training convention rather than validity of the convention. Step 2 (fitted input called prediction): Section 4.4 uses the resulting scorer on LLM responses to produce the benchmark's 'quantitative metric' (Section 3.3) and its Table 3 model rankings, which are statistically forced by the fitted labels. No external anchor exists anywhere in the paper: no inter-annotator agreement for the 18 category scores, no independent expert rating of the 50-prompt responses, no comparison against an external benchmark. The Section 4.3.3 human panel only prefers model variant A over B on high/low-error samples; it never validates the target scores. The Limitations section candidly states the mapping 'lacks universal applicability and thus relies heavily on subjective judgment,' which confirms (rather than mitigates) the closed loop. Mitigating factors: there is no self-citation chain (the reference list contains no papers by the present authors), the training pipeline is reproducible (78,200 samples, seed 3407, released code), and the scorer genuinely generalizes within its own convention, so the derivation is not a pure algebraic identity. Score 6: the central claims, accurate relevance scoring and the quantitative LLM evaluation metric, reduce to the authors' fitted label convention, though the model training and prompt set retain independent engineering content.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

All training targets and all downstream scores are derived from Table 4's hand-set mapping and the alpha+beta=20 Beta sampling; no independent ground truth or external benchmark is introduced. The ledger is therefore dominated by free parameters rather than by measured or externally validated quantities.

free parameters (4)
  • Per-category hard relevance scores (18 categories) = Human-set values in Table 4, e.g., Gov 1.0, News 0.7, Law_A 0.9, Others 0.0
    These hand-assigned values define the ground truth for training, evaluation, and all downstream LLM scores. They are not measured or externally validated.
  • Beta concentration parameter (alpha+beta) = 20
    Chosen by hand in Section 4.2 to control variance of the Beta diffusion; every training target depends on this choice.
  • Boundary score adjustments = Government to ~0.95, Others to ~0.15
    Ad hoc smoothing applied before Beta diffusion, modifying the initial hard scores and therefore all generated soft labels.
  • Evaluation error tolerance thresholds = 0.1 and 0.2
    Used to convert continuous predictions into accuracy and F1 numbers in Table 1; changing these thresholds changes the headline results.
assumptions (5)
  • domain assumption Government-domain relevance is a well-defined scalar property of text, representable by a single number in [0,1].
    This is the foundation of the benchmark (Sections 3.1 and 4.2). No evidence is given that relevance is one-dimensional or independent of prompt context.
  • ad hoc to paper Beta diffusion around hand-set hard scores produces valid soft-label training targets.
    The noise-injection step in Section 4.2 changes every training target, but the paper does not compare the resulting soft scores to human relevance ratings.
  • domain assumption The 50 chosen prompts are representative of the space of government-domain LLM tasks.
    The prompt set is 45 author-written and 5 news-like prompts (Section 3.3), selected without external expert coverage analysis.
  • domain assumption ModernBERT's encoder is expressive enough to learn government-domain relevance from the constructed targets.
    ModernBERT is chosen mainly for speed and context length (Section 2.3, Appendix A.1); no comparison against other encoder architectures for this scoring task is provided.
  • standard math Beta distribution expectation and variance formulas support the diffusion design.
    Equation (4) and Figure 3 rely on the standard result that E[Beta(a,b)] = a/(a+b) and that variance is controlled by a+b.

how reviews work

0 comments
Cite this review

Pith. "Pith review of GovRelBench:A Benchmark for Government Domain Relevance." pith.science (2026). https://pith.science/paper/EYVDVWPL

@misc{pith2026250721419,
  author       = {Pith},
  title        = {Pith review of: GovRelBench:A Benchmark for Government Domain Relevance},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EYVDVWPL}},
  note         = {Machine review of arXiv:2507.21419}
}
read the original abstract

Current evaluations of LLMs in the government domain primarily focus on safety considerations in specific scenarios, while the assessment of the models' own core capabilities, particularly domain relevance, remains insufficient. To address this gap, we propose GovRelBench, a benchmark specifically designed for evaluating the core capabilities of LLMs in the government domain. GovRelBench consists of government domain prompts and a dedicated evaluation tool, GovRelBERT. During the training process of GovRelBERT, we introduce the SoftGovScore method: this method trains a model based on the ModernBERT architecture by converting hard labels to soft scores, enabling it to accurately compute the text's government domain relevance score. This work aims to enhance the capability evaluation framework for large models in the government domain, providing an effective tool for relevant research and practice. Our code and dataset are available at https://github.com/pan-xi/GovRelBench.

Figures

Figures reproduced from arXiv: 2507.21419 by the authors.

Figure 1
Figure 1. Overall pipeline of GovRelBench and GovRelBERT scoring workflow [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Two-step Construction of SoftGovScore: Category Mapping (Step 1) and Beta Diffusion (Step 2) [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Variance Profile of the Beta Distribution [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Reference Score Distribution fusion (Section 3.1). A consistency constraint of α + β = 20 was applied to all categories. This ensured uniform prior confidence while allowing the score variance to adjust appropriately based on the hard score—resulting in lower variance …

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

25 extracted references · 7 canonical work pages

  1. [1]

    Yuntao Bai, Saurav Kadavath, Sandipan Kundu, Amanda Askell, Jackson Kernion, Andy Jones, Anna Chen, Anna Goldie, Azalia Mirhoseini, Cameron McKinnon, et al. 2022. Constitutional ai: Harmlessness from ai feedback. arXiv preprint arXiv:2212.08073

    arXiv 2022

  2. [2]

    Jianlv Chen, Shitao Xiao, Peitian Zhang, Kun Luo, Defu Lian, and Zheng Liu. 2024. Bge m3-embedding: Multi-lingual, multi-functionality, multi-granularity text embeddings through self-knowledge distillation. arXiv preprint arXiv:2402.03216

    arXiv 2024

  3. [3]

    Katherine M Collins, Umang Bhatt, and Adrian Weller. 2022. Eliciting and learning with soft labels from every annotator. In Proceedings of the AAAI conference on human computation and crowdsourcing, volume 10, pages 40--52

    work page 2022

  4. [4]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 conference of the North American chapter of the association for computational linguistics: human language technologies, volume 1 (long and short papers), pages 4171--4186

    2019

  5. [5]

    hard” and “soft

    Aram Galstyan and Paul R Cohen. 2007. Empirical comparison of “hard” and “soft” label propagation for relational classification. In International Conference on Inductive Logic Programming, pages 98--111. Springer

    work page 2007

  6. [6]

    Pengcheng He, Xiaodong Liu, Jianfeng Gao, and Weizhu Chen. 2020. Deberta: Decoding-enhanced bert with disentangled attention. arXiv preprint arXiv:2006.03654

    arXiv 2020

  7. [7]

    Geoffrey Hinton, Oriol Vinyals, and Jeff Dean. 2015. Distilling the knowledge in a neural network. arXiv preprint arXiv:1503.02531

    arXiv 2015

  8. [8]

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, Weizhu Chen, et al. 2022. Lora: Low-rank adaptation of large language models. ICLR, 1(2):3

    2022

Show all 25 references
  1. [9]

    Jinqi Lai, Wensheng Gan, Jiayang Wu, Zhenlian Qi, and Philip S Yu. 2024. Large language models in law: A survey. AI Open

    2024

  2. [10]

    Zhenzhong Lan, Mingda Chen, Sebastian Goodman, Kevin Gimpel, Piyush Sharma, and Radu Soricut. 2019. Albert: A lite bert for self-supervised learning of language representations. arXiv preprint arXiv:1909.11942

    2019 arXiv

  3. [11]

    Jean Lee, Nicholas Stevens, and Soyeon Caren Han. 2025. Large language models in finance (finllms). Neural Computing and Applications, pages 1--15

    2025

  4. [12]

    Lei Liu, Xiaoyan Yang, Junchi Lei, Xiaoyang Liu, Yue Shen, Zhiqiang Zhang, Peng Wei, Jinjie Gu, Zhixuan Chu, Zhan Qin, et al. 2024. A survey on medical large language models: Technology, application, trustworthiness, and future directions. arXiv preprint arXiv:2406.03712

    2024 arXiv

  5. [13]

    Shuo Liu, Lin Zhang, Weidong Liu, Jianfeng Zhang, Donghui Gao, and Xiaofeng Jia. 2025. The evaluation framework and benchmark for large language models in the government affairs domain. ACM Transactions on Intelligent Systems and Technology

    2025

  6. [14]

    Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Mandar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692

    2019 arXiv

  7. [15]

    MOP-LIWU Community and MNBVC Team . 2023. Mnbvc: Massive never-ending bt vast chinese corpus. https://github.com/esbatmop/MNBVC

    2023

  8. [16]

    Quang Nguyen, Hamed Valizadegan, and Milos Hauskrecht. 2014. Learning classification models with soft-label information. Journal of the American Medical Informatics Association, 21(3):501--508

    2014

  9. [17]

    Long Ouyang, Jeffrey Wu, Xu Jiang, Diogo Almeida, Carroll Wainwright, Pamela Mishkin, Chong Zhang, Sandhini Agarwal, Katarina Slama, Alex Ray, et al. 2022. Training language models to follow instructions with human feedback. Advances in neural information processing systems, 3...

    2022

  10. [18]

    Noam Shazeer. 2020. Glu variants improve transformer. arXiv preprint arXiv:2002.05202

    2020 arXiv

  11. [19]

    Xiaofeng Shi, Lulu Zhao, Hua Zhou, and Donglin Hao. 2024. https://doi.org/10.57967/hf/3488 Industrycorpus2

    2024 doi

  12. [20]

    Nidhi Vyas, Shreyas Saxena, and Thomas Voice. 2020. Learning soft labels via meta learning. arXiv preprint arXiv:2009.09496

    2020 arXiv

  13. [21]

    Benjamin Warner, Antoine Chaffin, Benjamin Clavi \'e , Orion Weller, Oskar Hallstr \"o m, Said Taghadouini, Alexis Gallagher, Raja Biswas, Faisal Ladhak, Tom Aarsen, et al. 2024. Smarter, better, faster, longer: A modern bidirectional encoder for fast, memory efficient, and lo...

    2024 arXiv

  14. [22]

    Zhao Xue, Hanyu Zhao, Sha Yuan, and Yequan Wang. 2022. https://doi.org/10.57760/sciencedb.o00126.00004 WuDaoCorpora Text

    2022

  15. [23]

    Shengyu Zhang, Linfeng Dong, Xiaoya Li, Sen Zhang, Xiaofei Sun, Shuhe Wang, Jiwei Li, Runyi Hu, Tianwei Zhang, Fei Wu, et al. 2023. Instruction tuning for large language models: A survey. arXiv preprint arXiv:2308.10792

    2023

  16. [24]

    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...

  17. [25]

    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 6, 2026 · model on record in the stance chip above.