How Amazon is Rethinking human evaluation for generative large language models [Breakdowns]
artificialintelligencemadesimple.substack.com
Reducing Bias and Improving Consistency in LLM Human Evaluations