We are delighted to announce that our paper “TART: An Open-Source Tool-Augmented Framework for Explainable Table-based Reasoning” has been awarded Best Paper Runner-Up (Top 3 out of 59 submissions; Award Rate: 5.1%) at the 3rd Table Representation Learning Workshop @ NeurIPS 2024.
About the Paper
TART introduces an open-source tool-augmented framework designed to improve explainable table-based reasoning. By leveraging external tools to carry out arithmetic and logical operations over tables, TART enables models to generate transparent and verifiable reasoning steps. This allows for greater interpretability, robustness, and reliability compared to standard chain-of-thought reasoning approaches.
About the Workshop
The Table Representation Learning (TRL) Workshop at NeurIPS focuses on advancing methods for understanding, reasoning, and learning from structured tabular data. The workshop serves as a premier venue for researchers working at the intersection of NLP, ML, and structured data reasoning.
📅 Awarded on: December 14, 2024
📍 Conference: TRL Workshop @ NeurIPS 2024
✍️ Authors: Xinyuan Lu, Liangming Pan, Yubo Ma, Preslav Nakov, Min-Yen Kan