Saved in:
Bibliographic Details
Main Authors: Anvekar, Tejas, Park, Junha, Garimella, Aparna, Gupta, Vivek
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2512.15907
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911610012958720
author Anvekar, Tejas
Park, Junha
Garimella, Aparna
Gupta, Vivek
author_facet Anvekar, Tejas
Park, Junha
Garimella, Aparna
Gupta, Vivek
contents Evaluating the quality of tables generated by large language models (LLMs) remains an open challenge: existing metrics either flatten tables into text, ignoring structure, or rely on fixed references that limit generalization. We present TabReX, a reference-less, property-driven framework for evaluating tabular generation via graph-based reasoning. TabReX converts both source text and generated tables into canonical knowledge graphs, aligns them through an LLM-guided matching process, and computes interpretable, rubric-aware scores that quantify structural and factual fidelity. The resulting metric provides controllable trade-offs between sensitivity and specificity, yielding human-aligned judgments and cell-level error traces. To systematically asses metric robustness, we introduce TabReX-Bench, a large-scale benchmark spanning six domains and twelve planner-driven perturbation types across three difficulty tiers. Empirical results show that TabReX achieves the highest correlation with expert rankings, remains stable under harder perturbations, and enables fine-grained model-vs-prompt analysis establishing a new paradigm for trustworthy, explainable evaluation of structured generation systems.
format Preprint
id arxiv_https___arxiv_org_abs_2512_15907
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TabReX : Tabular Referenceless eXplainable Evaluation
Anvekar, Tejas
Park, Junha
Garimella, Aparna
Gupta, Vivek
Computation and Language
Evaluating the quality of tables generated by large language models (LLMs) remains an open challenge: existing metrics either flatten tables into text, ignoring structure, or rely on fixed references that limit generalization. We present TabReX, a reference-less, property-driven framework for evaluating tabular generation via graph-based reasoning. TabReX converts both source text and generated tables into canonical knowledge graphs, aligns them through an LLM-guided matching process, and computes interpretable, rubric-aware scores that quantify structural and factual fidelity. The resulting metric provides controllable trade-offs between sensitivity and specificity, yielding human-aligned judgments and cell-level error traces. To systematically asses metric robustness, we introduce TabReX-Bench, a large-scale benchmark spanning six domains and twelve planner-driven perturbation types across three difficulty tiers. Empirical results show that TabReX achieves the highest correlation with expert rankings, remains stable under harder perturbations, and enables fine-grained model-vs-prompt analysis establishing a new paradigm for trustworthy, explainable evaluation of structured generation systems.
title TabReX : Tabular Referenceless eXplainable Evaluation
topic Computation and Language
url https://arxiv.org/abs/2512.15907