diff options
| author | YurenHao0426 <Blackhao0426@gmail.com> | 2026-07-22 13:28:23 -0500 |
|---|---|---|
| committer | YurenHao0426 <Blackhao0426@gmail.com> | 2026-07-22 13:28:23 -0500 |
| commit | c27e0824a41e6c54dd9079f17c20c306c5f2e9c9 (patch) | |
| tree | 052d79f45fe1f8fc4e2de853fec6e592acdfed58 /THEORY.md | |
| parent | d4a521072a25d8afdd62021f43538d7dad7586fd (diff) | |
results: close hierarchical task-scalar calibration gate
Diffstat (limited to 'THEORY.md')
| -rw-r--r-- | THEORY.md | 6 |
1 files changed, 5 insertions, 1 deletions
@@ -247,7 +247,11 @@ The executable audit checks the complete BatchNorm-coupled antithetic JVP to predicted Q/R moments equal the exact causal delta-rule targets to `2.3e-16` relative error. Actual training never performs that copy. These identities establish mechanics only; a frozen-forward causal-capture gate must pass before -any task-scale endpoint is opened. +any task-scale endpoint is opened. That empirical V4 gate subsequently failed: +the stable rate left early alignment near zero, while larger rates caused a +58.5x feedback norm ratio or nonfinite state. The unbiased identity therefore +does not imply usable finite-budget signal in the 267,904-dimensional joint +parameter space. ## 2. What is sufficient for a descent step |
