plan(sp17): allocate ISV slots [474..483) for dueling-Q diagnostics

Browse Source

9 slots: A_var_ema × 4 branches (per-branch advantage variance EMA)
+ V_share × 4 branches (V/(V+A) magnitude share) + adaptive
advantage_clip_bound. All Pearl-A first-observation bootstrap with
fold-reset registry entries + dispatch arms in
trainer/training_loop.rs::reset_named_state. Bump ISV_TOTAL_DIM 474 →
483 + layout fingerprint seed.

Per `feedback_isv_for_adaptive_bounds`: advantage_clip_bound is
producer-tracked from p99(|A_centered|) × 1.5 safety factor, never
hardcoded. Bounds [0.1, 100.0] are Category-1 dimensional safety floors.

No consumer change in this commit (additive infrastructure). Phase 1
mean-centering producer (atomic across compute_expected_q +
c51_loss + c51_grad + mag_concat_qdir) lands in the next 4 commits.

Plan: docs/superpowers/plans/2026-05-08-sp17-dueling-q-network.md

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

...

This commit is contained in:

jgrusewski

2026-05-08 20:35:29 +02:00

parent 25ff1d4195

commit a225926e5f

5 changed files with 277 additions and 2 deletions

Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL

Download Patch File Download Diff File Expand all files Collapse all files

13

crates/ml/src/cuda_pipeline/gpu_dqn_trainer.rs

File diff suppressed because one or more lines are too long