results: record corrected scaling and residualization pilots

1 files changed, 74 insertions, 9 deletions

diff --git a/RESULTS.md b/RESULTS.md
index d2bdc48..d1a7c75 100644
--- a/RESULTS.md
+++ b/RESULTS.md
@@ -1,16 +1,40 @@
 # SDIL — results log (seed 0 unless noted; runs on timan107 GTX-1080 / ep_pascal)
 
+## 2026-07-21 audit and version boundary
+
+Git was initialized after inheriting the project. The original code/results are preserved at
+`215acff`. Four audit findings changed the canonical implementation:
+
+1. `49b86d6`: residual-block local updates now include the branch Jacobian
+   `res_alpha=1/sqrt(depth)`. The omission preserved cosine direction but inflated update norms by
+   `sqrt(depth)`, confounding cross-depth optimization comparisons.
+2. `5255166`: the Harnett baseline is now a per-neuron affine soma-dendrite fit
+   `ahat_l = p_l * h_l + b_l`, rather than a dense population map. This follows the paper's
+   per-cell least-squares somatic-event to dendritic-event relation.
+3. `c2769e6`: causal perturbation targets and predicted residuals are measured on the same
+   pre-update network state.
+4. `0f107d1` / `e0bfdfd`: simultaneous hidden-layer perturbations reduce calibration from
+   quadratic to linear depth work; batching directions makes the reduction visible in GPU wall
+   time. `66e3034` fixes diagnostics so raw-apical ablations report the signal actually used.
+
+All older `deepres_*` files predate item 1. Their alignment directions remain informative, but
+their accuracy and stability should not be pooled with the corrected runs below.
+
 ## Method
 SDIL (Somato-Dendritic Innovation Learning), a local non-backprop rule inspired by Harnett 2026.
 Per hidden layer l:
 - forward: `u_l = W_l h_{l-1}`, `h_l = φ(u_l)`
 - apical feedback `a_l = A_l c` (c = output error e = softmax−onehot, broadcast)
-- predictor `ĥ_l = P_l h_l`; teaching signal (innovation) `r_l = a_l − ĥ_l`
+- default per-neuron predictor `ĥ_l = p_l ⊙ h_l + b_l`; teaching signal `r_l = a_l − ĥ_l`
 - three-factor update `ΔW_l = η (r_l ⊙ φ'(u_l)) h_{l-1}^T`
 - `A_l` LEARNED by amortized node perturbation (q_l ≈ −∇_{h_l}L, forward-only, no weight transport)
 - `P_l` learned on neutral (c=0) periods only, so it strips the soma-predictable nuisance
   without eating the teaching signal.
 
+The scalable perturbation mode injects independent Rademacher interventions at all hidden layers
+in the same antithetic forward evaluations. Cross-layer interference adds variance but averages to
+zero; it trains all `A_l` using `O(depth)` rather than `O(depth^2)` forward work.
+
 Distinct from the failed `~/sdrn` (fixed-random feedback + post-hoc residualization) and from
 Dual Prop / DFA / FA: the feedback pathway is *learned by causal perturbation*, so the residual is
 aligned by construction rather than being a random projection.
@@ -22,7 +46,45 @@ magnitude-invariant so alignment looked fine while optimization blew up. Fixing
 "worse than DFA everywhere" to "matches DFA on MNIST, beats it on FMNIST/CIFAR". (Codex independently
 flagged the same σ-scale issue.)
 
-## Accuracy (test %, seed 0)
+## Corrected narrow depth pilot (CIFAR-10, width 64, 5 epochs, seed 0)
+
+| depth | method | test acc | early-third alignment | wall time |
+|------:|--------|---------:|----------------------:|----------:|
+| 30 | BP | 41.31% | — | 29.4 s |
+| 30 | DFA | 37.31% | 0.084 | 19.4 s |
+| 30 | SDIL layerwise | 42.01% | 0.371 | 339.5 s |
+| 30 | SDIL multiplexed+batched | 41.46% | 0.370 | 47.3 s |
+| 60 | BP | 42.93% | — | 63.7 s |
+| 60 | DFA | 37.31% | 0.041 | 41.1 s |
+| 60 | SDIL layerwise | 40.79% | 0.328 | 1238.5 s |
+| 60 | SDIL multiplexed+batched | 41.10% | 0.332 | 90.3 s |
+
+Narrowing the model exposed the performance difference that width-256 flattened CIFAR had hidden.
+DFA alignment halves from depth 30 to 60, while both SDIL estimators remain near 0.33–0.37. Batched
+multiplexing preserves alignment while accelerating SDIL 7.2x at d30 and 13.7x at d60. Its wall
+growth from d30 to d60 is 1.91x, versus 3.65x for layerwise calibration. These are pilot, single-seed
+results; the ~0.5-point layerwise/multiplexed accuracy difference is not yet statistically resolved.
+
+## Harnett-specific residualization (MNIST, 15 epochs, seed 0)
+
+| nuisance rho | raw apical acc | residual acc | raw teaching alignment | residual alignment |
+|-------------:|---------------:|-------------:|-----------------------:|-------------------:|
+| 0.00 | 97.31% | 97.31% | 0.138 | 0.138 |
+| 0.02 | 96.90% | 97.27% | 0.096 | 0.156 |
+| 0.05 | 96.48% | 97.24% | 0.092 | 0.160 |
+| 0.10 | 95.87% | 97.32% | 0.068 | 0.140 |
+| 0.20 | 95.60% | 97.31% | 0.103 | 0.147 |
+| 0.50 | 8.92% | 97.18% | 0.000 | 0.149 |
+| 1.00 | 8.92% | 97.29% | 0.000 | 0.140 |
+| 2.00 | 9.82% | 97.31% | 0.000 | 0.121 |
+
+At zero nuisance, subtraction is a no-op. As normal soma-dendrite coupling grows, raw feedback
+degrades continuously and eventually enters a self-reinforcing Hebbian failure, while the learned
+innovation remains stable. The low-rho sweep is important: it shows a graded 0.37→1.71 point gap
+before the strong-contamination phase transition, rather than only an artificial chance-level
+contrast.
+
+## Legacy accuracy (pre-audit implementation)
 | task                         | BP        | DFA   | SDIL          |
 |------------------------------|-----------|-------|---------------|
 | MNIST  d3 / d5 / d7 / d10    | 98.3      | 97.3 / 97.2 / 97.1 / 96.8 | 97.4 / 97.3 / 97.1 / 97.1 |
@@ -33,14 +95,14 @@ SDIL matches DFA on easy MNIST and **beats DFA on FashionMNIST (+1–2 pts) and
 On CIFAR the SDIL−DFA gap GROWS with depth (+1.8 → +2.7 → +5.0 at d5/10/20): DFA degrades with depth,
 SDIL holds.
 
-## Credit assignment / depth utility — the mechanistic result
+## Legacy credit assignment / depth utility
 Per-layer alignment cos(r_l, −∇_{h_l}L), early-third average, CIFAR no-BN residual:
 | depth | DFA early-align | SDIL early-align |
 |-------|-----------------|------------------|
 | 5     | 0.55            | 0.83             |
 | 10    | 0.36            | 0.86             |
 | 20    | 0.25            | 0.88             |
-| 30    | 0.19            | (running)        |
+| 30    | 0.19            | 0.87             |
 
 DFA's teaching signal to early/deep layers collapses toward random (~0.05–0.19) as depth grows;
 SDIL stays ~0.85 and is depth-invariant. This is the "does credit reach early layers" metric that
@@ -61,8 +123,11 @@ Analysis: `experiments/analyze_depth.py {depth,deepres}` → tables + figures in
 Smoke/mechanism checks: `experiments/smoke.py`.
 
 ## Open items
-- Complete CIFAR SDIL d30 (running) + seeds 1,2 (running) for mean±std.
-- Codex-suggested boosts (not yet applied): separate normalized feedback direction from a learned
-  per-layer scalar gain; Adam-like second-moment on local updates; forward-only gradient-norm gain
-  calibration. Current results already beat DFA without these.
-- GRAPE (ICLR2026) read for DFA-failure regimes only; treated as unverified, not a baseline.
+- Repeat corrected width-64 d30/d60 and nuisance transition runs for seeds 1–4.
+- Compare multiplexed calibration at fixed loss-evaluation budgets and sweep directions (4/8/16/32).
+- Add a magnitude-matched or clipped raw-apical control to separate directional contamination from
+  simple update explosion at large nuisance.
+- Validate simultaneous perturbation on a genuinely depth-necessary compositional task and a CNN,
+  not only flattened CIFAR.
+- FA's residual-network implementation and the current EP prototype are not yet strong enough for a
+  publication-quality scaling comparison; do not overstate those baselines.