RoT: Enhancing Table Reasoning with Iterative Row-Wise Traversals

2025-05-21

AI Summary

To address severe content hallucination, high training costs, and the trade-off between reliability and efficiency in table reasoning, this paper proposes Row-wise Traversal (RoT), a training-free inference paradigm. RoT enables controllable expansion of reasoning paths through structured row-by-row traversal, reflective refinement, and chain-of-thought compression—achieving substantial robustness gains with zero training overhead. By leveraging large language models’ intrinsic reflective capabilities, RoT avoids excessive generation over tabular text, effectively mitigating content hallucination. Evaluated on WikiTableQuestions and TableBench, RoT achieves state-of-the-art performance, improving average accuracy by 4.3% over reinforcement-learning-based LLMs (RLLMs) while significantly reducing inference token consumption. The method thus delivers both high reliability and computational efficiency without requiring fine-tuning or additional training data.

Technology Category

• Knowledge Representation and Reasoning: Computational Complexity of Reasoning
• Cognitive Modeling & Cognitive Systems: Conceptual Inference and Reasoning
• Natural Language Processing: Safety and Robustness

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

The table reasoning task, crucial for efficient data acquisition, aims to answer questions based on the given table. Recently, reasoning large language models (RLLMs) with Long Chain-of-Thought (Long CoT) significantly enhance reasoning capabilities, leading to brilliant performance on table reasoning. However, Long CoT suffers from high cost for training and exhibits low reliability due to table content hallucinations. Therefore, we propose Row-of-Thought (RoT), which performs iteratively row-wise table traversal, allowing for reasoning extension and reflection-based refinement at each traversal. Scaling reasoning length by row-wise traversal and leveraging reflection capabilities of LLMs, RoT is training-free. The sequential traversal encourages greater attention to the table, thus reducing hallucinations. Experiments show that RoT, using non-reasoning models, outperforms RLLMs by an average of 4.3%, and achieves state-of-the-art results on WikiTableQuestions and TableBench with comparable models, proving its effectiveness. Also, RoT outperforms Long CoT with fewer reasoning tokens, indicating higher efficiency.

Problem

• Enhancing table reasoning with iterative row-wise traversals
• Reducing hallucinations in table content reasoning
• Achieving training-free efficient reasoning with fewer tokens

Innovation

• Row-wise iterative traversal for table reasoning
• Training-free method leveraging LLM reflection
• Reduces hallucinations with sequential attention