Solovay's theorems of arithmetical adequacy for GL — LessWrong