SimpleQA

SimpleQA is a benchmark containing 1,000 prompts used to evaluate the short-form factuality of Large Language Models (LLMs), addressing limitations in OpenAI's original SimpleQA dataset.

Papers using SimpleQA (5)

SimpleQA Verified: A Reliable Factuality Benchmark to Measure Parametric Knowledge2025 · 1 citesCo-LMLM: Continuous-Query Limited Memory Language Models2026PIXELRAG: Web Screenshots Beat Text for Retrieval-Augmented Generation2026EAPO: Enhancing Policy Optimization with On-Demand Expert Assistance2025Fusion Steering: Prompt-Specific Activation Control2025