RelationalFactQA: A Benchmark for Evaluating Tabular Fact Retrieval from Large Language Models
This addresses the challenge of factual accuracy in LLMs for structured knowledge retrieval, establishing a benchmark for future progress, but it is incremental as it builds on existing factuality evaluation efforts.
The paper tackled the problem of evaluating factuality in large language models (LLMs) for generating structured, multi-record tabular outputs, and found that even state-of-the-art LLMs struggle significantly, not exceeding 25% factual accuracy, with performance degrading as output dimensionality increases.
Factuality in Large Language Models (LLMs) is a persistent challenge. Current benchmarks often assess short factual answers, overlooking the critical ability to generate structured, multi-record tabular outputs from parametric knowledge. We demonstrate that this relational fact retrieval is substantially more difficult than isolated point-wise queries, even when individual facts are known to the model, exposing distinct failure modes sensitive to output dimensionality (e.g., number of attributes or records). To systematically evaluate this under-explored capability, we introduce RelationalFactQA, a new benchmark featuring diverse natural language questions (paired with SQL) and gold-standard tabular answers, specifically designed to assess knowledge retrieval in a structured format. RelationalFactQA enables analysis across varying query complexities, output sizes, and data characteristics. Our experiments reveal that even state-of-the-art LLMs struggle significantly, not exceeding 25% factual accuracy in generating relational outputs, with performance notably degrading as output dimensionality increases. These findings underscore critical limitations in current LLMs' ability to synthesize structured factual knowledge and establish RelationalFactQA as a crucial resource for measuring future progress in LLM factuality.