Tabular Representation, Noisy Operators, and Impacts on Table Structure Understanding Tasks in LLMs
This work addresses the robustness of LLMs in handling messy tabular data, which is incremental as it builds on prior research by introducing noise operations.
The study investigated how different tabular representations and noise operations affect LLMs' performance on structural understanding tasks, finding that noise operations significantly impact performance across formats.
Large language models (LLMs) are increasingly applied for tabular tasks using in-context learning. The prompt representation for a table may play a role in the LLMs ability to process the table. Inspired by prior work, we generate a collection of self-supervised structural tasks (e.g. navigate to a cell and row; transpose the table) and evaluate the performance differences when using 8 formats. In contrast to past work, we introduce 8 noise operations inspired by real-world messy data and adversarial inputs, and show that such operations can impact LLM performance across formats for different structural understanding tasks.