AI Paper Atlas

Research map / Large language models / Reasoning and chain of thought

TaREx: Reinforcement Learning for Code-Driven Table Reasoning

Fangyu Lei, Jinxiang Meng, Yiming Huang, Shizhu He, Jun Zhao, Kang Liu

AAAI 2026 · Large language models › Reasoning and chain of thought · approach: RL with verifiable rewards for reasoning · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Proposes TaREx, a reinforcement-learning approach that trains models to solve table reasoning tasks by generating code, targeting structure comprehension, hallucination-prone text reasoning, and multi-step processing over tables.

Problem

Table reasoning is hard due to diverse hierarchical table structures, reliance on complex logical and numerical reasoning that makes text-only methods hallucinate, and the need for multi-step processing over multiple and lengthy tables.

Approach

Uses reinforcement learning to drive code-based table reasoning, so the model executes code over tables rather than reasoning purely in text. Details beyond the abstract are not available.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use DataFrame observation instead of raw text (ablation)Table-RL · WikiTQscore: +1.15 (81.00 vs. 79.03)Improved
use DataFrame observation instead of raw text (ablation)Table-RL · FinQAscore: +1.70 (69.61 vs. 67.91)Improved
use DataFrame observation instead of raw text (ablation)Table-RL · MultiHierttscore: +1.17 (43.52 vs. 42.35)Improved
add hybrid code-and-text action space (code execution) (ablation)DF-Table-RL · TabFactaccuracy: +2.19 (90.76 vs. 88.57)Improved
add hybrid code-and-text action space (code execution) (ablation)DF-Table-RL · FinQAscore: +1.42 (78.97 vs. 77.55)Improved
add multi-turn interaction (ablation)DF-Table-Code-RL · WikiTQscore: 84.12 vs. 81.00Improved
add multi-turn interaction (ablation)DF-Table-Code-RL · TabFactaccuracy: 92.35 vs. 90.76Improved
add multi-turn interaction (ablation)DF-Table-Code-RL · FinQAscore: 69.81 vs 69.61Improved
add difficulty-based data filtering (ablation)TaREx-7B · WikiTQscore: +1.16 (83.62 vs 82.46)Improved
add difficulty-based data filtering (ablation)TaREx-7B · MultiHierttscore: +1.49 (47.82 vs 46.33)Improved
add compact filtering of timeout and max-step trajectories (ablation)TaREx · training dynamicsresponse lengthImproved
add trajectory filtering of stagnant turns (ablation)TaREx · training dynamicsgradient varianceImproved

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