Pith. sign in

REVIEW 9 cited by

TAPAS: Weakly Supervised Table Parsing via Pre-training

Not yet reviewed by Pith; the record is open.

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

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2004.02349 v2 pith:7DBMCEWV submitted 2020-04-05 cs.IR cs.AIcs.CLcs.LG

Jonathan Herzig , Pawe{l} Krzysztof Nowak , Thomas Müller , Francesco Piccinno , Julian Martin Eisenschlos This is my paper
classification cs.IRcs.AIcs.CLcs.LG
keywords tapasformslogicalparsingsemantictablesstate-of-the-artsupervision
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Answering natural language questions over tables is usually seen as a semantic parsing task. To alleviate the collection cost of full logical forms, one popular approach focuses on weak supervision consisting of denotations instead of logical forms. However, training semantic parsers from weak supervision poses difficulties, and in addition, the generated logical forms are only used as an intermediate step prior to retrieving the denotation. In this paper, we present TAPAS, an approach to question answering over tables without generating logical forms. TAPAS trains from weak supervision, and predicts the denotation by selecting table cells and optionally applying a corresponding aggregation operator to such selection. TAPAS extends BERT's architecture to encode tables as input, initializes from an effective joint pre-training of text segments and tables crawled from Wikipedia, and is trained end-to-end. We experiment with three different semantic parsing datasets, and find that TAPAS outperforms or rivals semantic parsing models by improving state-of-the-art accuracy on SQA from 55.1 to 67.2 and performing on par with the state-of-the-art on WIKISQL and WIKITQ, but with a simpler model architecture. We additionally find that transfer learning, which is trivial in our setting, from WIKISQL to WIKITQ, yields 48.7 accuracy, 4.2 points above the state-of-the-art.

Discussion (0). Sign in to comment.

Forward citations

Cited by 9 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. TEN: Table Explicitization, Neurosymbolically

    cs.CL 2025-08 unverdicted novelty 6.0 of 10

    A neurosymbolic system with structural decomposition prompting and a checker-driven self-debug loop improves table extraction from semistructured text over purely neural baselines.

  2. TableMind: An Autonomous Programmatic Agent for Tool-Augmented Table Reasoning

    cs.AI 2025-09 conditional novelty 5.0 of 10

    TableMind, a two-stage SFT-plus-RL agent trained on an 8B model, reports state-of-the-art results on three table reasoning benchmarks.

  3. ExpliCIT-QA: Explainable Code-Based Image Table Question Answering

    cs.CL 2025-07 conditional novelty 5.0 of 10

    ExpliCIT-QA turns table-image question answering into a traceable sequence of table extraction, natural-language reasoning, code generation, execution, and explanation.

  4. Multimodal Tabular Reasoning with Privileged Structured Information

    cs.LG 2025-06 conditional novelty 5.0 of 10

    An 8B multimodal LLM trained on 9k reasoning traces distilled from structured tables reaches state-of-the-art open-source accuracy on table-image question answering and fact verification.

  5. Ta-G-T: Subjectivity Capture in Table to Text Generation via RDF Graphs

    cs.CL 2025-07 conditional novelty 4.0 of 10

    A modular T5-based pipeline using RDF triples, sentence aggregation, and style transfer generates factual text with subjective interpretations from tables, achieving moderate gains over several LLM baselines.

  6. Advancing Retrieval-Augmented Generation for Structured Enterprise and Internal Data

    cs.CL 2025-07 reject novelty 4.0 of 10

    An enterprise RAG framework combining hybrid retrieval, cross-encoder reranking, and structure-aware table indexing claims relative gains of 15% in Precision@5, 13% in Recall@5, and 16% in MRR over a dense-only baseline.

  7. Interactive Text-to-SQL via Expected Information Gain for Disambiguation

    cs.DB 2025-07 reject novelty 4.0 of 10

    An interactive text-to-SQL framework selects clarification questions by expected information gain over a distribution of candidate SQL queries.

  8. OpenTable-R1: A Reinforcement Learning Augmented Tool Agent for Open-Domain Table Question Answering

    cs.CL 2025-07 conditional novelty 4.0 of 10

    A 4B-parameter model fine-tuned with supervised cold-start and an asynchronous GRPO reinforcement learning variant reaches 86.2% exact match on a held-out subset of Open WikiTable by using search and SQL tools.

  9. LLM-Assisted Question-Answering on Technical Documents Using Structured Data-Aware Retrieval Augmented Generation

    cs.CL 2025-06 reject novelty 4.0 of 10

    A RAG pipeline with OCR, table and image to text conversion, and a RAFT-tuned reranker reports high QA scores, but its 50-question evaluation overlaps with its training manuals and its baseline comparison uses only 5 ...

Pith tools