AI.info

Hannah Rose Kirk

Explore Hannah Rose Kirk on AI.info.

• Reward Models Inherit Value Biases from Pretraining
• Measuring what Matters: Construct Validity in Large Language Model Benchmarks