Pith. sign in

REVIEW 3 major objections 5 minor 59 references

ProgramTab teaches LLMs to clean messy web tables with Python before extracting answers via SQL, outperforming other prompting methods on table QA and fact verification.

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-4.5

2026-07-14 06:16 UTC pith:WK4XO7IF

load-bearing objection Clean engineering pipeline that puts Python normalization before CoT-SQL; consistent gains and large-table robustness, but the fixed top-K=10 retrieval that drives code/SQL generation is under-tested. the 3 major comments →

arxiv 2607.11207 v1 pith:WK4XO7IF submitted 2026-07-13 cs.CL cs.AI

ProgramTab: Boosting Table Reasoning of LLMs via Programmatic Paradigm

Pei Guo , Enjie Liu , Yunzhi Tan , Mochi Gao , Jianxin Zhang , Ruichao Zhong , Juntao Li , Bo Hu
show 1 more author
Zang Li
This is my paper
classification cs.CL cs.AI
keywords table reasoninglarge language modelsin-context learningPython code generationtext-to-SQLtable preprocessingWikiTQTabFact
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.

Large language models struggle with table reasoning when tables are large or inconsistently formatted: long inputs hit context limits, and raw web tables often lack the structure and types needed for reliable SQL. ProgramTab addresses this by guiding an LLM, via few-shot prompts, through five steps: retrieve the most relevant rows with embeddings, select useful columns, generate Python code that standardizes formats and types and optionally adds derived columns, produce multi-step SQL with chain-of-thought, then read the SQL result to answer the question. On WikiTQ and TabFact the method beats prior LLM baselines under the same models, and it remains strong even when tables exceed typical context windows. The practical payoff is a length-agnostic pipeline that turns messy tables into clean, queryable data without extra training.

Core claim

By inserting an explicit Python preprocessing stage before SQL generation, and by decomposing questions into sub-queries, LLMs can handle large, inconsistently structured tables far more reliably than methods that feed whole tables or generate SQL alone; the resulting ProgramTab pipeline outperforms all compared LLM-based baselines on WikiTQ and TabFact while using only four LLM calls per example.

What carries the argument

ProgramTab: a five-stage in-context pipeline (row retrieval by embedding, column extraction, Python data-redefinition code, CoT SQL generation, answer synthesis) that first normalizes table cells then extracts the answer.

Load-bearing premise

The top-10 rows retrieved by a frozen embedding model, plus the columns chosen in one LLM pass, still contain the information needed to answer most questions.

What would settle it

On a held-out set of large tables where the gold answer rows rank outside the top-10 by the same embedding model, measure whether ProgramTab accuracy collapses relative to an oracle that is given the true relevant rows.

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

Share X Bluesky LinkedIn Reddit HN

If this is right

  • Table-reasoning systems can drop full-table context and still improve accuracy by first emitting executable Python normalizers.
  • Existing text-to-SQL pipelines become more robust on web tables once a short code-generation step unifies formats and types.
  • Latency and token cost stay low because only four greedy LLM samples are required per question.
  • The same normalize-then-query pattern extends immediately to hierarchical tables after a lightweight header-flattening pass.

Where Pith is reading between the lines

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

  • If row-retrieval coverage is the main failure mode, replacing the frozen embedder with a learned or iterative retriever should lift the remaining errors more than stronger code models alone.
  • The same Python-first normalization idea could be reused for semi-structured documents (HTML lists, CSV dumps) that currently break pure SQL agents.
  • Error analysis suggests code and SQL generation quality scale with model strength; weaker open models may need self-repair loops the paper currently limits to five retries.

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

3 major / 5 minor

Summary. The paper introduces ProgramTab, a five-stage in-context learning pipeline for LLM table reasoning: (1) embedding-based top-K row retrieval (GTE-base, K=10), (2) LLM column extraction, (3) LLM-generated Python code that standardizes formats, casts types, and optionally derives new columns, (4) CoT-decomposed text-to-SQL executed on the cleaned table, and (5) final answer generation from the SQL result. It targets two known failure modes—context-length limits on large tables and the structural inconsistency of web tables that makes pure SQL brittle. Experiments on WikiTQ, TabFact and (appendix) HiTab with GPT-3.5-Turbo, Llama-3.1-70B-Instruct and GPT-4o-mini report state-of-the-art numbers among prompting methods (e.g., 76.0 % denotation accuracy on WikiTQ and 89.6 % on TabFact with GPT-4o-mini), strong large-table and token-cutoff results, ablations isolating the Python-redefinition and CoT-SQL stages, efficiency counts, and an error taxonomy.

Significance. If the empirical claims hold, ProgramTab supplies a practical, training-free recipe that simultaneously mitigates length constraints and data-quality problems that currently limit LLM table reasoners. The explicit separation of Python normalization from SQL extraction is a clean engineering contribution; the efficiency analysis (only four LLM calls, ~4 s/sample) and the large-table stress tests (Table 4) are concrete strengths that go beyond many concurrent prompting papers. The work is therefore of clear interest to the table-reasoning and LLM-tool-use communities, provided the retrieval assumption that underpins the length-robustness claims is more thoroughly validated.

major comments (3)
  1. [§3.1, Appendix A.4, Table 4] §3.1 and Appendix A.4 state that top-K=10 GTE-base rows are used as the sole exemplars for column selection, Python redefinition and CoT-SQL generation, while the full (column-filtered) table is used only at execution time. The sole supporting statistic is that the gold answer appears among those 10 rows for 86 % of WikiTQ examples. No ablation replaces top-K with random, bottom-K or oracle rows, nor is final accuracy stratified by whether the answer (or the format variants / rare filter values needed for correct code) is present in the exemplars. Because the large-table gains (Table 4) and the headline “outperforms all LLM-based baselines” rest on this untested proxy, the central length-robustness claim remains incompletely supported.
  2. [§5.1, Table 3, Figure 4] Table 3 shows a 10.9-point drop on WikiTQ when data redefinition is removed, confirming that the Python stage is load-bearing. Yet the paper never reports how often the generated Python actually alters column types or extracts new columns, nor whether those alterations are correct on the full table versus the top-K exemplars. Without such diagnostics it is hard to know whether the observed gains come from genuine normalization or from incidental side-effects of the few-shot prompt.
  3. [§4.3, Figure 7] Implementation details (§4.3) state that after five failed retries on syntax errors the pipeline simply aborts. The fraction of such hard failures, and whether they are counted as incorrect answers, is never reported. Given that code and SQL errors dominate the manual error analysis (Figure 7), the missing failure-rate statistic leaves open the possibility that a non-negligible portion of the reported accuracy gap versus baselines is an artifact of the retry policy rather than of better reasoning.
minor comments (5)
  1. [Abstract, §1] Abstract and §1 use the phrase “momentous contents extraction”; “key” or “salient” would be clearer.
  2. [Figure 6] Figure 6 prompt contains the typo “venus” for “venue”; the same figure also mixes “SQL Answer” and “A:” labels inconsistently.
  3. [Tables 1–2] Table 1 and Table 2 mix GPT-4, Codex and GPT-4o-mini baselines; a short note clarifying that each ProgramTab row is compared only against methods that used the same (or weaker) backbone would avoid any appearance of unfair comparison.
  4. [Appendix A.5] Appendix A.5 HiTab numbers are given only for GPT-4o-mini; reporting the same three backbones used in the main tables would strengthen the hierarchical-table claim.
  5. [§4.3, Appendix A.2] The hyper-parameter table (temperature 0.6, top_p=1, K=10, five retries) is buried in Appendix A.2; moving the key values into the main experimental section would improve reproducibility.

Circularity Check

0 steps flagged

No circularity: empirical prompting pipeline evaluated on public benchmarks with fixed hyperparameters; no derivation reduces to fitted inputs or self-citation.

full rationale

ProgramTab is a systems/methods paper that composes known techniques (embedding-based row retrieval with frozen GTE-base, LLM column selection, few-shot Python normalization, CoT text-to-SQL, final LLM answer generation) and reports denotation/accuracy numbers on WikiTQ, TabFact, and HiTab against external baselines. Hyper-parameters (K=10, temperature 0.6, max five retries) are chosen once and held fixed; they do not appear as free parameters inside the reported accuracies. There is no mathematical derivation, uniqueness theorem, or fitted quantity that is later re-labeled a prediction. Related-work citations (Binder, Dater, NormTab, TabSQLify, Chain-of-Table, etc.) are to independent prior systems; none of the load-bearing claims rest on a self-citation of an unverified uniqueness result by the same authors. Ablations (w/o DR, w/o CoT SG) and large-table subsets are ordinary empirical controls, not circular reductions. The paper is therefore self-contained against external benchmarks; circularity score is zero.

Axiom & Free-Parameter Ledger

3 free parameters · 3 axioms · 1 invented entities

The central accuracy claims rest on a small set of free hyper-parameters, standard LLM in-context-learning assumptions, and the invented five-stage pipeline itself. No deep mathematical axioms are required; the work is empirical.

free parameters (3)
  • K (number of retrieved rows) = 10
    Fixed to 10 for all models; chosen by context-length considerations rather than cross-validation. Directly determines which table content the later LLM stages ever see.
  • LLM temperature / top_p / sample count = 0.6 / 1 / 1
    Temperature 0.6, top_p=1, n=1 for every stage; not swept.
  • max retries on code/SQL syntax error = 5
    Hard-coded to five; after five failures the example is abandoned.
axioms (3)
  • domain assumption Frozen embedding similarity (GTE-base) ranks rows sufficiently well that the top-10 contain the answer for most questions.
    Stated in §3.1 and quantified only post-hoc in Appendix A.4 (86 % coverage).
  • domain assumption In-context learning with a handful of hand-written examples is sufficient for an LLM to emit correct Python normalizers and CoT SQL.
    Core operating premise of every stage; never proved, only demonstrated empirically.
  • ad hoc to paper Web-table cells can be made SQL-ready by a short Python function that only uses string extraction, type casts and simple arithmetic.
    Assumed in the data-redefinition prompt (§3.3); fails on more complex hierarchical or multi-value cells (acknowledged in Limitations).
invented entities (1)
  • ProgramTab five-stage pipeline no independent evidence
    purpose: Orchestrates row retrieval, column extraction, Python redefinition, CoT SQL and final answer generation into a single length-robust table-reasoning method.
    The pipeline itself is the paper’s main contribution; it has no independent existence outside this work.

pith-pipeline@v1.1.0-grok45 · 20747 in / 2617 out tokens · 30880 ms · 2026-07-14T06:16:13.373648+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of ProgramTab: Boosting Table Reasoning of LLMs via Programmatic Paradigm." pith.science (2026). https://pith.science/paper/WK4XO7IF

@misc{pith2026260711207,
  author       = {Pith},
  title        = {Pith review of: ProgramTab: Boosting Table Reasoning of LLMs via Programmatic Paradigm},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WK4XO7IF}},
  note         = {Machine review of arXiv:2607.11207}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Table-based reasoning with large language models (LLMs), which requires reasoning based on natural language questions and structured tabular data, has gained widespread attention. However, a series of issues still constrain the application of this task. The previous approaches suffered from significant performance degradation when faced with large tables due to the difficulty of long text modeling and the limitation of input length for LLMs. The text-to-SQL approach is used to efficiently extract key information from tables and generate smaller sub-tables. However, tabular data, especially web tables, often lack the necessary structure and consistency, making them unsuitable for performing mathematical logic operations using SQL queries. We propose the ProgramTab framework, which guides LLMs employing in-context learning to perform tabular data preprocessing with Python code, as well as the momentous contents extraction with row and column extraction and SQL generation. The experiment results on table reasoning datasets demonstrate that the ProgramTab framework effectively deals with table-based reasoning tasks and outperforms all LLM-based baselines.

Figures

Figures reproduced from arXiv: 2607.11207 by Bo Hu, Enjie Liu, Jianxin Zhang, Juntao Li, Mochi Gao, Pei Guo, Ruichao Zhong, Yunzhi Tan, Zang Li.

Figure 1
Figure 1. Figure 1: An example of a table in WikiTQ dataset. [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The overview of ProgramTab for table-based reasoning. [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Prompt for LLMs to extract columns. of the cells to form a text segment, and an embed￾ding model is utilized to calculate the relevance score between the row data and the question. Ulti￾mately, the top K most relevant rows are selected as instances in the prompt templates of the following steps, which effectively alleviates the whole table as the input context. The value of ‘K’ is determined based on the c… view at source ↗
Figure 4
Figure 4. Figure 4: The prompt for LLMs to perform data redefi [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Prompt for LLMs to generate final answers. [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Statistic of different error types on TabFact [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 9
Figure 9. Figure 9: Comparison of the average number of table [PITH_FULL_IMAGE:figures/full_fig_p012_9.png] view at source ↗
Figure 8
Figure 8. Figure 8: Prompt for LLMs to generate SQL w/o CoT. [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗

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

59 extracted references · 10 linked inside Pith

  1. [1]

    Aho and Jeffrey D

    Alfred V. Aho and Jeffrey D. Ullman , title =. 1972

    1972

  2. [2]

    Publications Manual , year = "1983", publisher =

    1983

  3. [3]

    Chandra and Dexter C

    Ashok K. Chandra and Dexter C. Kozen and Larry J. Stockmeyer , year = "1981", title =. doi:10.1145/322234.322243

    doi:10.1145/322234.322243 1981

  4. [4]

    Scalable training of

    Andrew, Galen and Gao, Jianfeng , booktitle=. Scalable training of

  5. [5]

    Dan Gusfield , title =. 1997

    1997

  6. [6]

    Tetreault , title =

    Mohammad Sadegh Rasooli and Joel R. Tetreault , title =. Computing Research Repository , volume =. 2015 , url =

    2015

  7. [7]

    A Framework for Learning Predictive Structures from Multiple Tasks and Unlabeled Data , Volume =

    Ando, Rie Kubota and Zhang, Tong , Issn =. A Framework for Learning Predictive Structures from Multiple Tasks and Unlabeled Data , Volume =. Journal of Machine Learning Research , Month = dec, Numpages =

  8. [8]

    Md Mahadi Hasan Nahid and Davood Rafiei , booktitle=. Tab

  9. [9]

    International Conference on Learning Representations (ICLR) , year=

    Binding Language Models in Symbolic Languages , author=. International Conference on Learning Representations (ICLR) , year=

  10. [10]

    Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval , year=

    Yunhu Ye and Binyuan Hui and Min Yang and Binhua Li and Fei Huang and Yongbin Li , title=. Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval , year=

  11. [11]

    International Conference on Learning Representations (ICLR) , year=

    Chain-of-Table: Evolving Tables in the Reasoning Chain for Table Understanding , author=. International Conference on Learning Representations (ICLR) , year=

  12. [12]

    International Conference on Learning Representations (ICLR) , year=

    Tabfact: A large-scale dataset for table-based fact verification , author=. International Conference on Learning Representations (ICLR) , year=

  13. [13]

    2021 , journal=

    Feverous: Fact extraction and verification over unstructured and structured information , author=. 2021 , journal=

    2021

  14. [14]

    Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , publisher=

    Vivek Gupta and Maitrey Mehta and Pegah Nokhiz and Vivek Srikumar , title=. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , publisher=. 2020 , pages=

    2020

  15. [15]

    S truct GPT : A General Framework for Large Language Model to Reason over Structured Data

    Jiang, Jinhao and Zhou, Kun and Dong, Zican and Ye, Keming and Zhao, Xin and Wen, Ji-Rong. S truct GPT : A General Framework for Large Language Model to Reason over Structured Data. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 2023

    2023

  16. [16]

    Compositional Semantic Parsing on Semi-Structured Tables

    Pasupat Panupong and Liang Percy. Compositional Semantic Parsing on Semi-Structured Tables. Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). 2015

    2015

  17. [17]

    Explanatory and actionable debugging for machine learning: A tableqa demonstration

    Minseok Cho and Gyeongbok Lee and Seung-won Hwang. Explanatory and actionable debugging for machine learning: A tableqa demonstration. Proceedings of the 42nd International ACM SIGIR Conference on Research and Development in Information Retrieval. 2019

    2019

  18. [18]

    , author=

    Language models are few-shot learners. , author=. Advances in neural information processing systems , pages=

  19. [19]

    , author=

    Training Compute-Optimal Large Language Models. , author=. Preprint arXiv:2203.15556 , year=

    Pith/arXiv arXiv

  20. [20]

    , author=

    GPT-3.5-Turbo. , author=. Technical Report , year=

  21. [21]

    , author=

    GPT-4. , author=. Technical Report , year=

  22. [22]

    , author=

    LLaMA: Open and Efficient Foundation Language Models. , author=. Preprint arXiv:2302.13971 , year=

    Pith/arXiv arXiv

  23. [23]

    Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , publisher=

    TaPas: Weakly supervised table parsing via pre-training , author=. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics , publisher=

  24. [24]

    Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies , publisher=

    OmniTab: Pretraining with natural and synthetic data for few-shot table-based question answering , author=. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies , publisher=

    2022

  25. [25]

    STAR : SQL Guided Pre-Training for Context-dependent Text-to- SQL Parsing

    Cai, Zefeng and Li, Xiangyu and Hui, Binyuan and Yang, Min and Li, Bowen and Li, Binhua and Cao, Zheng and Li, Weijie and Huang, Fei and Si, Luo and Li, Yongbin. STAR : SQL Guided Pre-Training for Context-dependent Text-to- SQL Parsing. Findings of the Association for Computational Linguistics: EMNLP 2022. 2022

    2022

  26. [26]

    Large Language Models are few(1)-shot Table Reasoners

    Chen, Wenhu. Large Language Models are few(1)-shot Table Reasoners. Findings of the Association for Computational Linguistics: EACL 2023. 2023

    2023

  27. [27]

    ReAcTable: Enhancing ReAct for Table Question Answering

    Yunjia Zhang and Jordan Henkel and Avrilia Floratou and Joyce Cahoon. ReAcTable: Enhancing ReAct for Table Question Answering. Proceedings of the VLDB Endowment 17(8). 2024

    2024

  28. [28]

    Chain-of-thought prompting elicits reasoning in large language models

    Jason Wei and Xuezhi Wang and Dale Schuurmans and Maarten Bosma and Brian Ichter and Fei Xia and Ed Chi and Quoc Le and Denny Zhou. Chain-of-thought prompting elicits reasoning in large language models. Proceedings of the 36th International Conference on Neural Information Processing Systems. 2023

    2023

  29. [29]

    Preprint arXiv:2308.03281 , year=

    Towards General Text Embeddings with Multi-stage Contrastive Learning , author=. Preprint arXiv:2308.03281 , year=

    Pith/arXiv arXiv

  30. [30]

    TAPEX: Table pre-training via learning a neural SQL executor

    Qian Liu and Bei Chen and Jiaqi Guo and Morteza Ziyadi and Zeqi Lin and Weizhu Chen and Jian-Guang Lou. TAPEX: Table pre-training via learning a neural SQL executor. In International Conference on Learning Representations. 2022

    2022

  31. [31]

    R eas TAP : Injecting Table Reasoning Skills During Pre-training via Synthetic Reasoning Examples

    Yilun Zhao and Linyong Nan and Zhenting Qi and Rui Zhang and Dragomir Radev. R eas TAP : Injecting Table Reasoning Skills During Pre-training via Synthetic Reasoning Examples. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. 2022

    2022

  32. [32]

    O mni T ab: Pretraining with Natural and Synthetic Data for Few-shot Table-based Question Answering

    Zhengbao Jiang and Yi Mao and Pengcheng He and Graham Neubig and Weizhu Chen. O mni T ab: Pretraining with Natural and Synthetic Data for Few-shot Table-based Question Answering. Proceedings of the 2022 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2022

    2022

  33. [33]

    Colin Raffel and Noam Shazeer and Adam Roberts and Katherine Lee and Sharan Narang and Michael Matena and Yanqi Zhou and Wei Li and Peter J. Liu. Exploring the limits of transfer learning with a unified text-totext transformer. Journal of Machine Learning Research. 2020

    2020

  34. [34]

    Preprint arXiv:2210.11416 , year=

    Scaling instruction-finetuned language models , author=. Preprint arXiv:2210.11416 , year=

    Pith/arXiv arXiv

  35. [35]

    Lever: Learning to verify language-to-code generation with execution

    Ansong Ni and Srini Iyer and Dragomir Radev and Ves Stoyanov and Wen-tau Yih and Sida I Wang and Xi Victoria Lin. Lever: Learning to verify language-to-code generation with execution. Proceedings of the 40th International Conference on Machine Learning (ICML’23). 2023

    2023

  36. [36]

    Exploring Chain of Thought Style Prompting for Text-to- SQL

    Chang-Yu Tai and Ziru Chen and Tianshu Zhang and Xiang Deng and Huan Sun. Exploring Chain of Thought Style Prompting for Text-to- SQL. Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing. 2023

    2023

  37. [37]

    T a P as: Weakly Supervised Table Parsing via Pre-training

    Herzig, Jonathan and Nowak, Pawel Krzysztof and M. T a P as: Weakly Supervised Table Parsing via Pre-training. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 2020

    2020

  38. [38]

    Table Fact Verification with Structure-Aware Transformer

    Zhang, Hongzhi and Wang, Yingyao and Wang, Sirui and Cao, Xuezhi and Zhang, Fuzheng and Wang, Zhongyuan. Table Fact Verification with Structure-Aware Transformer. Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP). 2020

    2020

  39. [39]

    PASTA : Table-Operations Aware Fact Verification via Sentence-Table Cloze Pre-training

    Gu, Zihui and Fan, Ju and Tang, Nan and Nakov, Preslav and Zhao, Xiaoman and Du, Xiaoyong. PASTA : Table-Operations Aware Fact Verification via Sentence-Table Cloze Pre-training. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. 2022

    2022

  40. [40]

    Tao Yu and Chien-Sheng Wu and Xi Victoria Lin and bailin wang and Yi Chern Tan and Xinyi Yang and Dragomir Radev and richard socher and Caiming Xiong , booktitle=. Gra

  41. [41]

    T a C ube: Pre-computing Data Cubes for Answering Numerical-Reasoning Questions over Tabular Data

    Zhou, Fan and Hu, Mengkang and Dong, Haoyu and Cheng, Zhoujun and Cheng, Fan and Han, Shi and Zhang, Dongmei. T a C ube: Pre-computing Data Cubes for Answering Numerical-Reasoning Questions over Tabular Data. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. 2022

    2022

  42. [42]

    An Inner Table Retriever for Robust Table Question Answering

    Lin, Weizhe and Blloshmi, Rexhina and Byrne, Bill and de Gispert, Adria and Iglesias, Gonzalo. An Inner Table Retriever for Robust Table Question Answering. Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2023

    2023

  43. [43]

    BERT : Pre-training of Deep Bidirectional Transformers for Language Understanding

    Devlin, Jacob and Chang, Ming-Wei and Lee, Kenton and Toutanova, Kristina. BERT : Pre-training of Deep Bidirectional Transformers for Language Understanding. Proceedings of the 2019 Conference of the North A merican Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers). 2019

    2019

  44. [44]

    BART : Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension

    Lewis, Mike and Liu, Yinhan and Goyal, Naman and Ghazvininejad, Marjan and Mohamed, Abdelrahman and Levy, Omer and Stoyanov, Veselin and Zettlemoyer, Luke. BART : Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. 2020

    2020

  45. [45]

    Pengcheng He and Jianfeng Gao and Weizhu Chen , journal=. De

  46. [46]

    Advances in International Conference on Learning Representations , year=

    Complexity-based prompting for multi-step reasoning , author=. Advances in International Conference on Learning Representations , year=

  47. [47]

    Advances in International Conference on Learning Representations , year=

    Automatic chain of thought prompting in large language models , author=. Advances in International Conference on Learning Representations , year=

  48. [48]

    Preprint arXiv:2203.02155 , year=

    Training language models to follow instructions with human feedback , author=. Preprint arXiv:2203.02155 , year=

    Pith/arXiv arXiv

  49. [49]

    Preprint arXiv:2107.03374 , year=

    Evaluating large language models trained on code , author=. Preprint arXiv:2107.03374 , year=

    Pith/arXiv arXiv

  50. [50]

    Preprint arXiv:2211.12588 , year=

    Program of thoughts prompting: Disentangling computation from reasoning for numerical reasoning tasks , author=. Preprint arXiv:2211.12588 , year=

    Pith/arXiv arXiv

  51. [51]

    In International Conference on Machine Learning , year=

    PAL: Program-aided language models , author=. In International Conference on Machine Learning , year=

  52. [52]

    Preprint arXiv:2204.00498 , year=

    Evaluating the text-to-sql capabilities of large language models , author=. Preprint arXiv:2204.00498 , year=

    Pith/arXiv arXiv

  53. [53]

    E^5 : Zero-shot Hierarchical Table Analysis using Augmented LLM s via Explain, Extract, Execute, Exhibit and Extrapolate

    Zhang, Zhehao and Gao, Yan and Lou, Jian-Guang. E^5 : Zero-shot Hierarchical Table Analysis using Augmented LLM s via Explain, Extract, Execute, Exhibit and Extrapolate. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2024

    2024

  54. [54]

    Preprint arXiv:2407.03061 , year=

    ALTER: Augmentation for Large-Table-Based Reasoning , author=. Preprint arXiv:2407.03061 , year=

    Pith/arXiv arXiv

  55. [55]

    The Twelfth International Conference on Learning Representations , year=

    CABINET: Content Relevance-based Noise Reduction for Table Question Answering , author=. The Twelfth International Conference on Learning Representations , year=

  56. [56]

    H i T ab: A Hierarchical Table Dataset for Question Answering and Natural Language Generation

    Zhoujun Cheng and Haoyu Dong and Zhiruo Wang and Ran Jia and Jiaqi Guo and Yan Gao and Shi Han and Jian-Guang Lou and Dongmei Zhang. H i T ab: A Hierarchical Table Dataset for Question Answering and Natural Language Generation. Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). 2022

    2022

  57. [57]

    N orm T ab: Improving Symbolic Reasoning in LLM s Through Tabular Data Normalization

    Md Mahadi Hasan Nahid and Davood Rafiei. N orm T ab: Improving Symbolic Reasoning in LLM s Through Tabular Data Normalization. Findings of the Association for Computational Linguistics: EMNLP 2024. 2024

    2024

  58. [58]

    Preprint arXiv:2407.05952 , year=

    H-STAR: LLM-driven Hybrid SQL-Text Adaptive Reasoning on Tables , author=. Preprint arXiv:2407.05952 , year=

    Pith/arXiv arXiv

  59. [59]

    Shunyu Yao and Jeffrey Zhao and Dian Yu and Nan Du and Izhak Shafran and Karthik Narasimhan and Yuan Cao , booktitle =