Skip to main content
Jaehong Oh

∑ULR

The ULR Empirical Programme — From Observation to M1–M8

3 equations extracted from this document. Each equation pairs with the prose paragraph that immediately precedes it in the source — clicking the title above opens the full document.

Open source document →

  1. #1

    The observation of latent alignment and the need for a distinct ULR object are different claims. The empirical programme was designed to close the following chain:

    phenomenon→null robustness→gauge typing→strong baseline comparison→held-out behaviour→nonreduction verdict.\text{phenomenon} \to \text{null robustness} \to \text{gauge typing} \to \text{strong baseline comparison} \to \text{held-out behaviour} \to \text{nonreduction verdict}.
  2. #2

    Therefore,

    M8 verdict=NO\boxed{\text{M8 verdict}=\textbf{NO}}
  3. #3

    Even after M8, the observer-relative role problem could still be characterised exactly. For role-conditioned transcript laws , the equal-prior Bayes risk is

    R∗=1−∥PL−PI∥TV2R^*=\frac{1-\lVert P_L-P_I\rVert_{\mathrm{TV}}}{2}