AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

Same Content, Different Representations: A Controlled Study for Table QA

Yue Zhang, Seiji Maekawa, Nikita Bhutani

ICLR 2026 · Large language models › Retrieval and knowledge · approach: Explanatory and multiple-choice QA benchmarks · 5 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
represent table semi-structured (verbalized) instead of structuredXiYan · RePairTQA all splitsaccuracy (%): 69.55 -> 24.08Hurt
represent table semi-structured instead of structuredLLM-NL2SQL · RePairTQA all splitsaccuracy (%): 69.14 -> 38.65Hurt
represent table semi-structured instead of structuredGPT-4o · RePairTQA all splitsaccuracy (%): 45.37 -> 41.93Hurt
represent table semi-structured instead of structuredWeaver · RePairTQA all splitsaccuracy (%): 62.19 -> 57.70Hurt
use long tables instead of shortGPT-4o · RePairTQAaccuracy (%): 53.26 -> 27.98Hurt
use long tables instead of shortH-STAR · RePairTQAaccuracy (%): 53.59 -> 12.19Hurt
use noisy (incomplete) schema instead of cleanLLM-NL2SQL · RePairTQA lookupaccuracy (%): 74.74 -> 20.12Hurt
use noisy schema instead of cleanH-STAR · RePairTQA lookupaccuracy (%): 77.90 -> 84.91Improved
verbalize only query-related columns (ablation)GPT-4o · RePairTQA S3 semi-structuredaccuracy (%): 43.26 -> 46.45Improved
verbalize more columns (1 -> 6 random) (ablation)GPT-4o · RePairTQA S3 semi-structuredaccuracy (%): 44.68 -> 42.20Hurt
widen tables (narrow to wide)GPT-4o · BIRD-based splits structuredaccuracy (%): 52.75 -> 27.91Hurt
replace GPT-4o with Gemini-2.5-ProLLM · RePairTQA S1structured vs semi-structured gap: 92.63 -> 88.42No clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy