# CSM_RH Paper 26 v0.1

Campaign 25 localizes the fully distinct four-point fixed-power loss.

- The refined centered Hardy-Littlewood model has four-point scale H^2 log^2 H, so the model is not the H^4 barrier.
- D4HEG reduces to a bound on the aggregate actual-prime correlation error.
- Lambda-1=(Lambda-Lambda-sharp)+(Lambda-sharp-1) yields an exact 16-sector split.
- Current Lambda residual discorrelation is only H log^{-A}X, while d_k residuals receive H X^{-c}.
- The proof-level difference is W_Lambda=polylog versus W_dk=polynomial.
- The Lambda W-bound comes from Vinogradov-Korobov zero-free-region technology.
- D4HEG remains useful only if three-shift aggregation creates cancellation beyond current one-interval residual theory.
- Campaign 26 attacks that possibility.

No proof or disproof of RH is claimed.