AI.info

Christopher Summerfield

Explore Christopher Summerfield on AI.info.

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