How to mitigate sandbagging — LessWrong
lesswrong.com · saved by 1 readers
TL;DR: Fine-tuning access, data quality, and scorability are three main factors influencing the difficulty of mitigating sandbagging. In the figure b…
TL;DR: Fine-tuning access, data quality, and scorability are three main factors influencing the difficulty of mitigating sandbagging. In the figure b…