OPENAI'S NEW HALLUCINATION BENCHMARK

Summary from the TLDR AI newsletter

OpenAI has released the SimpleQA benchmark, which measures models' abilities around simple factual questions.