Pith. sign in

REVIEW 3 major objections 5 minor 40 references

KBAlign: Efficient Self Adaptation on Specific Knowledge Bases

T0 review · 3 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read KBAlign shows that a model adapted to a knowledge base using only its own self-annotated questions and answers captures about 90 percent of the performance gain of GPT-4-supervised fine-tuning, at a fraction of the cost.

desk verdict A practical self-supervised KB adaptation recipe with honest ablations, but the 90% headline rests on a single dataset and the self-generated training signal is never directly validated. read the letter →

arxiv 2411.14790 v4 pith:LFT2KGQ5 submitted 2024-11-22 cs.CL cs.AI

Zheni Zeng , Yuxuan Chen , Shi Yu , Ruobing Wang , Yukun Yan , Zhenghao Liu , Shuo Wang , Xu Han
show 2 more authors
Zhiyuan Liu Maosong Sun
This is my paper · ORCID
classification cs.CLcs.AI
keywords retrieval-augmentedgenerationknowledgebaseadaptationself-supervisedfine-tuningself-annotationiterativeself-verificationknowledge-basedquestionansweringparameter-efficienttuninglong-contextQA
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

KBAlign adapts a language model to a specific textual knowledge base using only the model's own capacity, with no human labels, no external model, and no supervised data. The paper claims this fully self-supervised recipe recovers about 90 percent of the performance gain obtained by fine-tuning on GPT-4-annotated data, at a fraction of the cost. On the LooGLE fact-QA benchmark, MiniCPM-2B rises from 30.92 to 54.09 F1 (versus 56.92 with GPT-4 data), while LLaMA-3.1-8B rises from 40.46 to 62.07 (versus 64.97). The two mechanisms are multi-grained self-annotation, which turns KB passages into question-answer pairs at short and long dependency scales, and iterative self-verification tuning, which lets the model check its own stage-by-stage predictions to converge faster. If correct, this means small and private knowledge bases can be exploited without the cost, confidentiality, or dependency problems of external supervision.

What carries the argument

The central machinery is the KBAlign loop of multi-grained self-annotation and iterative self-verify tuning. Multi-grained annotation operates at two scales: short-dependency annotation splits the KB into fixed-length chunks and generates direct question-answer pairs per chunk; long-dependency annotation concatenates related short segments so the model must produce answers that integrate information from several places. Iterative self-verify tuning splits the annotated data in parts, fine-tunes on one part, uses the resulting model to generate RAG predictions on the next part, and then has the model compare its prediction with the self-annotated answer to produce a verification judgment; the verification outputs are used as additional training targets alongside the QA pairs. This two-part loop is what lets a small model adapt to a KB without external signals, and it is the object the paper's efficiency and 90%-gain claims are about.

What would settle it

Test KBAlign on a domain where the backbone model is known to lack knowledge (e.g., a constructed corpus of novel facts); if the self-annotated answers are mostly wrong and the adapted model's downstream accuracy drops below the vanilla RAG baseline, the claim that self-annotation alone suffices would be overturned.

Watch

Extended reading notes

Core claim

The paper's discovery is that a model's own self-annotation, organized at multiple grains and combined with iterative self-verification, is a sufficient training signal to align a generative model with a small textual knowledge base for RAG-based question answering. The backbone model reads gold passages, generates questions, retrieves related context, and answers from the combined material; the resulting question-answer pairs are used to fine-tune the model. Long-dependency annotation concatenates related segments so that questions and answers require multi-hop integration, while short-dependency annotation covers single-paragraph facts. The annotated data is then split; the model is first tuned on one part, then asked to verify its own RAG prediction against the annotated answer on the next part, and the verification outputs are mixed into the next round of tuning. The paper reports that this pipeline yields about 90% of the performance gain of GPT-4-annotated fine-tuning on LooGLE, with larger gains on fact-focused tasks and smaller gains on tasks whose knowledge the backbone already largely possesses (such as Wikipedia-based ASQA), and that it generalizes across a 2B and an 8B backbone and across legal, biomedical, and long-form QA.

Load-bearing premise

The self-annotated questions and answers produced by the backbone model itself are accurate enough to serve as training targets, since KBAlign has no mechanism to filter out incorrect self-annotations.

Editorial extensions

If this is right

  • A 2B-parameter model adapted by KBAlign surpasses an 8B-parameter model and GPT-4o on the LooGLE fact-QA benchmark (54.09 vs 40.46 and 40.20 F1).
  • On ASQA and BioASQ the gains are smaller but still positive, indicating the method helps most when the KB content is not already mastered by the backbone model.
  • Removing the iterative self-verify step slows convergence; the verification task is what lets the model reach the same F1 with less training data.
  • Mixing a small amount of general instruction data during adaptation preserves most of the model's MMLU score while retaining most of the downstream improvement.
  • For practical deployment, annotating more self-generated data and tuning for a single epoch outperforms reusing a smaller set for multiple epochs.

Reading between the lines

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

  • Because Section 4.5 concedes that self-annotated data 'contains some bias or error,' a natural next step is to filter self-annotations by the model's own verification confidence; a threshold could close part of the remaining 10% gap to GPT-4-supervised data.
  • The method is likely most effective when the KB sits near the model's pretraining distribution; for genuinely novel factual content, the loop has no external check and may reinforce the model's own hallucinations, suggesting a hybrid design with occasional human verification.
  • A natural testable extension is to use the model's verification outputs as pseudo relevance labels for retriever adaptation, a direction the paper explicitly leaves open.
  • The query-expansion-by-prediction step suggests a cheaper variant: only expand the search query when the model's self-verification confidence is low, saving retrieval cost when the model is already certain.
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 proposes KBAlign, a self-supervised framework for adapting an LLM to a small textual knowledge base (KB) for retrieval-augmented generation. The method generates multi-grained self-annotated QA pairs (short-dependency and long-dependency), iteratively tunes the model with self-verification data built from its own predictions and self-annotated answers, and at inference applies query expansion. Experiments on LooGLE, ASQA, JEC-QA, and BioASQ with MiniCPM-2B and LLaMA-3.1-8B-Instruct report consistent gains over vanilla RAG on most metrics, and the paper claims that on LooGLE KBAlign achieves about 90% of the performance gain obtained with GPT-4-annotated training data. Detailed ablations include removing test-relevant knowledge, removing verification, and removing long-dependency annotation.

Significance. If the central claim is validated, KBAlign would be practically valuable because it removes the need for human annotation or external large-model supervision when adapting small models to specialized KBs. The paper has several strengths: it evaluates on four datasets in three domains, uses two backbone models of different sizes, includes a thoughtful 'w/o know' ablation that partially controls for answer leakage, and releases data and models. The iterative self-verification idea is sensible and the efficiency numbers are attractive. However, the headline 90% figure is currently supported mainly by LooGLE F1 after hyperparameters are selected using that same benchmark, and the quality of the self-annotated training signal is never directly measured. These issues make the central quantitative claim weaker than the abstract suggests, though the qualitative finding that self-annotation helps RAG adaptation appears defensible after revision.

major comments (3)
  1. [§4.4 (Numerical Analysis; Figs. 3–4)] The reported LooGLE numbers are selected, not predicted. The text states that the authors 'search the best values for key settings including the training steps, amount of data and iteration by evaluating checkpoints in process,' and Figures 3–4 use LooGLE F1 to choose the data density (15 items per 10,000 tokens) and the number of iterations. Because no held-out validation split is described, the headline comparison to GPT-4-supervised adaptation on LooGLE is made after tuning hyperparameters on the test benchmark itself. This is load-bearing for the central 90% claim: the same configuration is then reported in Tables 1 and 2. Please either fix hyperparameters on a separate validation set, report results across a range of settings without test-set selection, or explicitly label the reported numbers as development-set-selected and re-state the claim accordingly.
  2. [§3.3, Algorithm 1 (lines 28–32)] The self-verification signal is not an independent check on self-annotation quality. The verification target V_i is generated by the model from (Q_i, P_i, A_i), where A_i is the same self-annotated answer used as the golden answer in the QA loss terms of L_i. Any systematic error in the self-annotated answers is therefore propagated through both the QA loss and the verification loss, rather than being corrected. Section 4.5 explicitly concedes that 'the self-annotated contains some bias or error, and this may damage the model performance on related questions.' The 'w/o know' ablation in Table 2 removes test-relevant content from the annotation context, but it does not measure the semantic correctness of the self-generated QA pairs. Please add a direct accuracy check on a sample of self-annotated Q/A pairs (e.g., against golden answers or human ratings) and analyze how annotation noise affects the verification stage.
  3. [§4.4, Table 1] The abstract's claim that KBAlign 'significantly improves downstream QA accuracy across multiple domains' is not supported by all reported metrics. On JEC-QA multiple-choice, both backbones decline relative to vanilla RAG (MiniCPM-2B: 13.87 to 9.94; LLaMA-3.1-8B: 24.66 to 14.13), and on ASQA the gains are modest (Match +3.77 and +5.02 for the two models). Please qualify the domain-general claim, report significance testing across the three seeds, or provide an aggregated measure that justifies the word 'significantly.'
minor comments (5)
  1. [Introduction, §3.2, §5] There are several typos: 'purpose' should be 'propose', 'practictible' should be 'practicable', 'refering' should be 'referring', 'targetd' should be 'targeted', and 'summarizatio' should be 'summarization.'
  2. [Algorithm 1 and §3.3] The loss notation E[||M(Q)-A||] is not standard for autoregressive language-model training; please define the concatenation operator ⊕ and clarify that these are token-level cross-entropy losses, not Euclidean norms.
  3. [Table 1] The MiniCPM JEC-QA single-choice score is written as '13,87'; this should be '13.87' for consistency with the other numeric entries.
  4. [Table 2] The ablation row 'w/overify' is ambiguous; it denotes the setting without verification and should be written as 'w/o verify' to match the notation used elsewhere, such as 'w/oQE' and 'w/oRAG.'
  5. [§4.2 and §4.4] The 'LLM' metric uses GPT-4o as the judge, but no human-agreement or sample-validity check is reported; small differences such as +0.92 on BioASQ in Table 1 should be interpreted with caution. Also, the 'GPT Data' ablation in Table 2 uses GPT-4-turbo, while RAFT in §4.1 uses GPT-4o; please clarify which annotation model underlies the 'GPT-4-supervised' comparison in the abstract.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: KBAlign's central gains are validated on external QA benchmarks, and the self-annotation/self-verification loop is a bootstrap training signal rather than a derivation that reduces to its own inputs.

full rationale

KBAlign's claimed result — reaching about 90% of the GPT-4-supervised gain with self-annotation — is an empirical comparison. The training targets (A, V in Algorithm 1) are generated by the backbone model itself, but the evaluation numbers in Tables 1-3 come from external QA benchmarks (LooGLE, ASQA, JEC-QA, BioASQ) measured against vanilla RAG and a GPT-4-annotated baseline; no reported metric is defined as a function of the self-generated labels or training loss. The self-verification step (Algorithm 1 lines 28-31) uses the model's own prediction P_i and self-annotated answer A_i to synthesize verification text V_i, and then trains on <Q_i⊕P_i, V_i>. This is a self-referential training loop, but it is not a circular derivation: the paper does not claim to predict the benchmark from the loop, and the w/o verify and w/o know ablations provide external checks that part of the gain is attributable to task format and general KB knowledge rather than leakage of exact test answers. The Section 4.5 admission that 'the self-annotated contains some bias or error' is a limitation of the training-signal quality, not a definitional equivalence. The numerical analysis selecting data density and iteration count by evaluating checkpoints on LooGLE F1 is a test-set-selection weakness, but it does not make the central claim equivalent to its inputs by construction. Self-citations (e.g., JEC-QA, Ding et al. instruction data) are dataset/tool references and are not load-bearing. Consequently no specific circular reduction can be quoted, and the circularity score is 0.

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

The central claim rests on several empirical premises: the model's self-annotations are accurate enough to train on, the retriever is adequate, the four datasets represent the target scenario, and the GPT-4o judge is a fair evaluator. Hyperparameters, including data density, iterations, and training epochs, are selected by observing benchmark performance, which means the reported numbers are partly tuned to the test tasks.

free parameters (5)
  • Self-annotation data density = 15 QA pairs per 10,000 tokens for LooGLE; increased for ASQA
    Chosen by evaluating checkpoints on the LooGLE test set (Section 4.4, Fig. 3). Affects the 90% claim and all reported numbers.
  • Verify data proportion = 25% verify, 75% QA
    Fixed in experiments (Section 3.3) without ablation across values; may not be optimal across domains.
  • Number of iterations = 2-3 iterations
    Recommended at least 3 iterations after tuning on LooGLE (Section 4.4, Fig. 4).
  • Training epochs = 1 epoch on full data; 2 epochs on half data hurts
    Selected based on Fig. 4 analysis; no validation split described.
  • Learning rate, batch size, LoRA rank = LR 1e-5, batch 8 for MiniCPM; LoRA alpha 16, rank 8 for LLaMA
    Grid search near MiniCPM sample code values (Appendix A.2). Standard but still benchmark-dependent.
assumptions (4)
  • domain assumption The backbone LLM can faithfully generate question-answer pairs from given KB text (self-annotation reliability).
    The entire training signal comes from M(Cg) and M(Q, C) in Algorithm 1; if the model cannot extract accurate info, the fine-tuning targets are corrupt. The paper admits annotation errors in Section 4.5.
  • domain assumption The retriever R returns relevant enough contexts during annotation and inference.
    Retrieval quality strongly influences QA performance; the paper mentions this in the Discussion and uses fixed bge retrievers.
  • domain assumption The benchmark datasets (LooGLE, ASQA, JEC-QA, BioASQ) are representative of small-scale KB adaptation scenarios.
    Generalization to other domains is inferred from these four datasets; no formal argument is provided.
  • domain assumption GPT-4o as an LLM judge reliably measures answer quality.
    The 'LLM' metric used in all tables is GPT-4o semantic judgment; the same model family is used for the GPT-4-supervised baseline, creating a potential evaluation bias.

how reviews work

0 comments
Cite this review

Pith. "Pith review of KBAlign: Efficient Self Adaptation on Specific Knowledge Bases." pith.science (2026). https://pith.science/paper/LFT2KGQ5

@misc{pith2026241114790,
  author       = {Pith},
  title        = {Pith review of: KBAlign: Efficient Self Adaptation on Specific Knowledge Bases},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LFT2KGQ5}},
  note         = {Machine review of arXiv:2411.14790}
}
read the original abstract

Although retrieval-augmented generation (RAG) remains essential for knowledge-based question answering (KBQA), current paradigms face critical challenges under specific domains. Existing methods struggle with targeted adaptation on small-scale KBs: vanilla unsupervised training exhibits poor effectiveness, while fine-tuning incurs prohibitive costs of external signals. We present KBAlign, a self-supervised framework that enhances RAG systems through efficient model adaptation. Our key insight is to leverage the model's intrinsic capabilities for knowledge alignment through two innovative mechanisms: multi-grained self-annotation that captures global knowledge for data construction, and iterative tuning that accelerates convergence through self verification. This framework enables cost-effective model adaptation to specific textual KBs, without human supervision or external model assistance. Experiments demonstrate that KBAlign can achieve 90\% of the performance gain obtained through GPT-4-supervised adaptation, while relying entirely on self-annotation of much smaller models. KBAlign significantly improves downstream QA accuracy across multiple domains with tiny costs, particularly benefiting scenarios requiring deep knowledge integration from specialized corpora. We release our experimental data, models, and process analyses to the community for further exploration (https://github.com/thunlp/KBAlign).

Figures

Figures reproduced from arXiv: 2411.14790 by the authors.

Figure 1
Figure 1. KBAlign schematic. We design special self [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. (a) Details for the KBAlign framework; (b) Instances for different annotation strategies and tasks. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. The impact of training amount on LooGLE performance. ‘w GPT‘ refers to training with GPT￾annotated data. models improve. Interestingly, directly learning without iterative tuning (dotted curve) also displays a similar trend, while the tipping point for slowing growth comes much later. This reveals the possible mechanism of self-verify task, that is, to guide the model to focus more on the problems of current stage, … view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The impact of iteration times and data amount [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Cases for KBAlign and baseline comparison. We display the translation for the Chinese JEC-QA task. [PITH_FULL_IMAGE:figures/full_fig_p008_5.png]
Figure 6
Figure 6. Figure 6: More cases for KBAlign and baseline comparison. [PITH_FULL_IMAGE:figures/full_fig_p013_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

40 extracted references · 10 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]

    Zixiang Chen, Yihe Deng, Huizhuo Yuan, Kaixuan Ji, and Quanquan Gu. 2024. Self-play fine-tuning converts weak language models to strong language models. arXiv preprint arXiv:2401.01335

    arXiv 2024

  3. [3]

    Daixuan Cheng, Shaohan Huang, and Furu Wei. 2023. Adapting large language models via reading comprehension. arXiv preprint arXiv:2309.09530

    arXiv 2023

  4. [4]

    Jiaxi Cui, Zongjian Li, Yang Yan, Bohua Chen, and Li Yuan. 2023. Chatlaw: Open-source legal large language model with integrated external knowledge bases. arXiv preprint arXiv:2306.16092

    arXiv 2023

  5. [5]

    Ning Ding, Yujia Qin, Guang Yang, Fuchao Wei, Zonghan Yang, Yusheng Su, Shengding Hu, Yulin Chen, Chi-Min Chan, Weize Chen, et al. 2022. Delta tuning: A comprehensive study of parameter efficient methods for pre-trained language models. arXiv preprint arXiv:2203.06904

    arXiv 2022

  6. [6]

    Qingxiu Dong, Lei Li, Damai Dai, Ce Zheng, Zhiyong Wu, Baobao Chang, Xu Sun, Jingjing Xu, and Zhifang Sui. 2022. A survey on in-context learning. arXiv preprint arXiv:2301.00234

    arXiv 2022

  7. [7]

    Tianyu Gao, Howard Yen, Jiatong Yu, and Danqi Chen. 2023. Enabling large language models to generate text with citations. In Empirical Methods in Natural Language Processing (EMNLP)

    work page 2023

  8. [8]

    Edward J Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu, Yuanzhi Li, Shean Wang, Lu Wang, and Weizhu Chen. 2021. Lora: Low-rank adaptation of large language models. arXiv preprint arXiv:2106.09685

    arXiv 2021

Show all 40 references
  1. [9]

    Shengding Hu, Yuge Tu, Xu Han, Chaoqun He, Ganqu Cui, Xiang Long, Zhi Zheng, Yewei Fang, Yuxiang Huang, Weilin Zhao, et al. 2024. Minicpm: Unveiling the potential of small language models with scalable training strategies. arXiv preprint arXiv:2404.06395

    2024 arXiv

  2. [10]

    Jiaxin Huang, Shixiang Gu, Le Hou, Yuexin Wu, Xuezhi Wang, Hongkun Yu, and Jiawei Han. 2023 a . Large language models can self-improve. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 1051--1068

    2023

  3. [11]

    Jie Huang, Xinyun Chen, Swaroop Mishra, Huaixiu Steven Zheng, Adams Wei Yu, Xinying Song, and Denny Zhou. 2023 b . Large language models cannot self-correct reasoning yet. In The Twelfth International Conference on Learning Representations

    2023

  4. [12]

    Kevin Maik Jablonka, Qianxiang Ai, Alexander Al-Feghali, Shruti Badhwar, Joshua D Bocarsly, Andres M Bran, Stefan Bringuier, L Catherine Brinson, Kamal Choudhary, Defne Circi, et al. 2023. 14 examples of how llms can transform materials science and chemistry: a reflection on a...

    2023

  5. [13]

    Xue Jiang, Yihong Dong, Lecheng Wang, Fang Zheng, Qiwei Shang, Ge Li, Zhi Jin, and Wenpin Jiao. 2023. Self-planning code generation with large language models. ACM Transactions on Software Engineering and Methodology

    2023

  6. [14]

    Qiao Jin, Yifan Yang, Qingyu Chen, and Zhiyong Lu. 2024. Genegpt: Augmenting large language models with domain tools for improved access to biomedical information. Bioinformatics, 40(2):btae075

    2024

  7. [15]

    Gonzalez, Hao Zhang, and Ion Stoica

    Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph E. Gonzalez, Hao Zhang, and Ion Stoica. 2023. Efficient memory management for large language model serving with pagedattention. In Proceedings of the ACM SIGOPS 29th Symposium on Operating S...

    2023

  8. [16]

    u ttler, Mike Lewis, Wen-tau Yih, Tim Rockt \

    Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich K \"u ttler, Mike Lewis, Wen-tau Yih, Tim Rockt \"a schel, et al. 2020. Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Proc...

    2020

  9. [17]

    Jiaqi Li, Mengmeng Wang, Zilong Zheng, and Muhan Zhang. 2023. Loogle: Can long-context language models understand long contexts? arXiv preprint arXiv:2311.04939

    2023 arXiv

  10. [18]

    Xun Liang, Shichao Song, Zifan Zheng, Hanyu Wang, Qingchen Yu, Xunkai Li, Rong-Hua Li, Feiyu Xiong, and Zhiyu Li. 2024. Internal consistency and self-feedback in large language models: A survey. arXiv preprint arXiv:2407.14507

    2024 arXiv

  11. [19]

    Chin-Yew Lin. 2004. Rouge: A package for automatic evaluation of summaries. In Text summarization branches out, pages 74--81

    2004

  12. [20]

    Chen Ling, Xujiang Zhao, Jiaying Lu, Chengyuan Deng, Can Zheng, Junxiang Wang, Tanmoy Chowdhury, Yun Li, Hejie Cui, Xuchao Zhang, et al. 2023. Domain specialization as the key to make large language models disruptive: A comprehensive survey. arXiv preprint arXiv:2305.18703

    2023 arXiv

  13. [21]

    Ali Madani, Ben Krause, Eric R Greene, Subu Subramanian, Benjamin P Mohr, James M Holton, Jose Luis Olmos, Caiming Xiong, Zachary Z Sun, Richard Socher, et al. 2023. Large language models generate functional protein sequences across diverse families. Nature Biotechnology, 41(8...

    2023

  14. [22]

    Kishore Papineni, Salim Roukos, Todd Ward, and Wei-Jing Zhu. 2002. Bleu: a method for automatic evaluation of machine translation. In Proceedings of the 40th annual meeting of the Association for Computational Linguistics, pages 311--318

    2002

  15. [23]

    Hongjin Qian, Peitian Zhang, Zheng Liu, Kelong Mao, and Zhicheng Dou. 2024. Memorag: Moving towards next-gen rag via memory-inspired knowledge discovery. arXiv preprint arXiv:2409.05591

    2024 arXiv

  16. [24]

    Yujia Qin, Shengding Hu, Yankai Lin, Weize Chen Chen, Ning Ding, Ganqu Cui, Zheni Zeng, and Yufei Huang. 2023. Tool learning with foundation models. arXiv preprint arXiv:2304.08354

    2023 arXiv

  17. [25]

    Ivan Stelmakh, Yi Luan, Bhuwan Dhingra, and Ming-Wei Chang. 2022. Asqa: Factoid questions meet long-form answers. arXiv preprint arXiv:2204.06092

    2022 arXiv

  18. [26]

    Ruoxi Sun, Sercan \"O Arik, Alex Muzio, Lesly Miculicich, Satya Gundabathula, Pengcheng Yin, Hanjun Dai, Hootan Nakhost, Rajarishi Sinha, Zifeng Wang, et al. 2023. Sql-palm: Improved large language model adaptation for text-to-sql (extended). arXiv preprint arXiv:2306.00739

    2023 arXiv

  19. [27]

    Zhen Tan, Alimohammad Beigi, Song Wang, Ruocheng Guo, Amrita Bhattacharjee, Bohan Jiang, Mansooreh Karami, Jundong Li, Lu Cheng, and Huan Liu. 2024. Large language models for data annotation: A survey. arXiv preprint arXiv:2402.13446

    2024 arXiv

  20. [28]

    Zhen Wan, Yating Zhang, Yexiang Wang, Fei Cheng, and Sadao Kurohashi. 2024. Reformulating domain adaptation of large language models as adapt-retrieve-revise: A case study on chinese legal domain. In Findings of the Association for Computational Linguistics ACL 2024, pages 5030--5041

    2024

  21. [29]

    Benyou Wang, Qianqian Xie, Jiahuan Pei, Zhihong Chen, Prayag Tiwari, Zhao Li, and Jie Fu. 2023 a . Pre-trained language models in biomedical domain: A systematic survey. ACM Computing Surveys, 56(3):1--52

    2023

  22. [30]

    Liang Wang, Nan Yang, and Furu Wei. 2023 b . Query2doc: Query expansion with large language models. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 9414--9423

    2023

  23. [31]

    Shijie Wu, Ozan Irsoy, Steven Lu, Vadim Dabravolski, Mark Dredze, Sebastian Gehrmann, Prabhanjan Kambadur, David Rosenberg, and Gideon Mann. 2023. Bloomberggpt: A large language model for finance. arXiv preprint arXiv:2303.17564

    2023 arXiv

  24. [32]

    Shitao Xiao, Zheng Liu, Peitian Zhang, and Niklas Muennighof. 2023. C-pack: Packaged resources to advance general chinese embedding. arXiv preprint arXiv:2309.07597

    2023 arXiv

  25. [33]

    Ming Xu. 2023. https://github.com/shibing624/text2vec text2vec: A tool for text to vector . Software

    2023

  26. [34]

    Di Zhang, Wei Liu, Qian Tan, Jingdan Chen, Hang Yan, Yuliang Yan, Jiatong Li, Weiran Huang, Xiangyu Yue, Dongzhan Zhou, et al. 2024. Chemllm: A chemical large language model. arXiv preprint arXiv:2402.06852

    2024 arXiv

  27. [35]

    Weinberger, and Yoav Artzi

    Tianyi Zhang*, Varsha Kishore*, Felix Wu*, Kilian Q. Weinberger, and Yoav Artzi. 2020. Bertscore: Evaluating text generation with bert. In International Conference on Learning Representations

    2020

  28. [36]

    Zihan Zhao, Da Ma, Lu Chen, Liangtai Sun, Zihao Li, Hongshen Xu, Zichen Zhu, Su Zhu, Shuai Fan, Guodong Shen, et al. 2024. Chemdfm: Dialogue foundation model for chemistry. arXiv preprint arXiv:2401.14818

    2024 arXiv

  29. [37]

    Yaowei Zheng, Richong Zhang, Junhao Zhang, Yanhan Ye, Zheyan Luo, Zhangchi Feng, and Yongqiang Ma. 2024. http://arxiv.org/abs/2403.13372 Llamafactory: Unified efficient fine-tuning of 100+ language models . In Proceedings of the 62nd Annual Meeting of the Association for Compu...

    2024 arXiv

  30. [38]

    Haoxi Zhong, Chaojun Xiao, Cunchao Tu, Tianyang Zhang, Zhiyuan Liu, and Maosong Sun. 2020. Jec-qa: a legal-domain question answering dataset. In Proceedings of the AAAI conference on artificial intelligence, volume 34, pages 9701--9708

    2020

  31. [39]

    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...

  32. [40]

    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 12, 2026 · model on record in the stance chip above.