What to Keep and What to Drop: Adaptive Table Filtering Framework
This addresses a bottleneck for researchers and practitioners using LLMs for table-based reasoning, offering an incremental improvement through a modular filtering pipeline.
The paper tackles the problem of large language models struggling with large tables due to input length limits by proposing ATF, an adaptive filtering framework that reduces table cells by 70% and boosts performance on out-of-domain TableQA tasks, though with slight drops on Table Fact Verification.
Large language models (LLMs) for table-based reasoning often struggle with large tables due to input length limits. We propose ATF (Adaptive Table Filtering Framework), a modular and question-aware filtering pipeline that prunes uninformative columns and rows using LLM-generated column descriptions, clustering, and sparse-dense alignment scores. ATF integrates seamlessly with existing models (e.g., TAPAS, TAPEX) without retraining. Experiments show that ATF reduces table cells by 70%, boosting performance on out-of-domain TableQA tasks while causing slight performance drops on Table Fact Verification, where full-table context is more critical. These results highlight ATF's ability to adaptively balance informativeness and minimalism across tasks. Our code available at: https://github.com/torijune/ATF-Adaptive-Table-Filtering-Framework