six general QA

The 'six general QA' dataset/benchmark evaluates the performance of models in answering general knowledge questions across various domains.

Papers using six general QA (1)

s3: You Don't Need That Much Data to Train a Search Agent via RL2025