docs: map frozen evidence to paper narrative

1 files changed, 176 insertions, 0 deletions

diff --git a/PAPER_PLAN.md b/PAPER_PLAN.md
new file mode 100644
index 0000000..a401909
--- /dev/null
+++ b/PAPER_PLAN.md
@@ -0,0 +1,176 @@
+# ICLR 2027 paper plan
+
+This document maps the frozen evidence to a defensible paper narrative. It is
+not permission to promote development results or to hide failed gates.
+
+## One-sentence claim
+
+Per-neuron subtraction of the apical component predictable from neutral
+somatic activity isolates an innovation signal that can preserve useful local
+credit when raw feedback is contaminated by ordinary soma-coupled traffic;
+when paired with causally learned feedback, the resulting local learner
+preserves accuracy and alignment over depth substantially better than fixed
+direct feedback in the audited regime.
+
+This sentence deliberately says neither that cortex implements
+backpropagation nor that the current model explains Harnett's temporal-error
+signature. “Scales on standard residual networks” may be added only after the
+untouched Oral-A confirmation passes.
+
+## Working title
+
+**Learning from the Unexpected: Somato-Dendritic Innovations for Local Credit
+Assignment**
+
+If Oral-A A4 passes cleanly, “Scalable” may be added to the title. If A3 or A4
+fails, the title stays mechanism-focused and the standard-scale result appears
+as a limitation rather than a title claim.
+
+## Abstract logic
+
+1. Biological and local-learning models often treat raw apical activity as a
+   teaching signal even though it also carries ordinary state/context traffic.
+2. Inspired by the per-neuron residual measured by Harnett and colleagues,
+   define teaching as apical activity minus its neutral-period prediction from
+   the same neuron's soma.
+3. Show that this is the orthogonal conditional innovation within the chosen
+   somatic function class; norm matching cannot reproduce its directional
+   effect. Calibrate the inherited feedback map using causal perturbations and
+   update forward weights with local eligibilities.
+4. Report the frozen residual-necessity panel: at strong predictable traffic,
+   innovation retains `97.35%` while raw and norm-matched raw reach about
+   chance, with an `87.038 +/- 0.607`-point paired gain over norm-matched raw.
+5. Report the frozen depth panel: SDIL changes by only
+   `-0.214 +/- 0.349` points over 12x hidden depth while DFA alignment falls to
+   `0.047`; qualify that flattened CIFAR is depth-flat.
+6. Insert the native-baseline and standard-ResNet result only after their
+   strict imports/gates complete. End with the supported scope: innovation is
+   useful for soma-predictable traffic, while learned feedback amortization
+   remains the useful-depth bottleneck.
+
+The abstract should not mention the failed desired-velocity hypothesis unless
+the paper explicitly positions that falsification as a contribution. It must
+not imply that no-traffic scaling establishes the necessity of residualization.
+
+## Contribution statements
+
+1. **Mechanism.** A per-cell neutral-period innovation rule for separating
+   task instruction from ordinary apical traffic, with raw and norm-matched
+   controls.
+2. **Theory.** A conditional-projection account of what residualization can
+   optimally remove, its failure boundary, a descent condition separating
+   direction/gain/curvature, and exact simultaneous-perturbation variance.
+   The conditional-expectation theorem itself is standard and is not claimed
+   as mathematical novelty.
+3. **Evidence.** Frozen, seed-complete tests of innovation necessity, depth
+   preservation, alignment, query/MAC/memory cost, and direct-versus-amortized
+   causal feedback.
+4. **Scientific negatives.** Broad top-down traffic, useful-depth learned
+   vectorization, and the Harnett desired-velocity signature fail their frozen
+   gates, defining the method's actual boundary.
+5. **Standard scale, conditional.** Add only if Oral-A passes its validation
+   and untouched confirmation gates.
+
+## Figure order in the manuscript
+
+The current filenames reflect generation history rather than final paper
+numbering.
+
+1. **Mechanism and necessity:** current `figure3_innovation`. Lead with the
+   subtraction diagram, accuracy under traffic, and used-signal alignment.
+2. **Accuracy/cost:** current `figure1_pareto`, augmented by the completed
+   native-author table in the text or supplement. Do not place unmatched
+   architectures on one purported equal-compute frontier.
+3. **Depth scaling:** current `figure2_scaling`. If Oral-A A4 passes, add or
+   replace with the ResNet20/32/56 accuracy, alignment-retention, and local-cost
+   panel; retain the MLP result as controlled diagnosis.
+4. **Mechanism/cost anatomy:** query-budget retention, direct-NP diagnosis,
+   predictor timescale, and failure boundaries. This can be a main figure if
+   space permits or a dense supplementary figure.
+5. **Biological signatures:** the complete failed Oral-B screen belongs in the
+   supplement/limitations unless the paper foregrounds falsification. Report
+   its positive decodability and negative causal-role results together.
+
+## Section skeleton
+
+### 1. Introduction
+
+- Mixed apical traffic is the unaddressed problem, not merely transporting an
+  output error to a dendrite.
+- Harnett motivates a per-neuron residual operation without proving that it is
+  a plasticity signal.
+- Existing dendritic, feedback-alignment, Dual Prop, BurstCCN, and learned
+  synthetic-feedback methods define the prior-art boundary.
+
+### 2. Somato-dendritic innovation learning
+
+- Forward dynamics and local eligibility.
+- Neutral predictor and innovation.
+- Causal perturbation calibration, clearly attributed to learned synthetic
+  feedback prior art.
+- Online-control variant described only as a tested hypothesis, not as the
+  successful method.
+
+### 3. Theory and resource accounting
+
+- Conditional projection and task-period absorption.
+- Hidden alignment versus parameter descent.
+- Simultaneous perturbation bias/variance.
+- Logical loss queries, MACs, memory, and wall time as separate axes.
+
+### 4. Does innovation matter?
+
+- Frozen 60-run predictable-traffic panel and matched-raw control.
+- Endogenous C1 near-miss/failure and FashionMNIST recovery failure.
+- The allowed claim is restricted to the predictor's somatic information.
+
+### 5. Does causal local credit survive depth?
+
+- Five-depth BP/FA/DFA/SDIL panel and local Pareto frontier.
+- Direct-NP useful-depth diagnosis and the failed learned-vectorizer branches.
+- Low-query confirmation.
+- Standard ResNet results only according to the frozen A1--A4 branch.
+
+### 6. Relation to biological signatures
+
+- Residual decorrelation, decoder, and plasticity-lesion positives.
+- Wrong sign-inversion, error-magnitude dominance, and weak acute online
+  lesion; desired velocity is unsupported.
+
+### 7. Limitations and discussion
+
+- Predictor observability and distribution invariance.
+- Perturbation variance and feedback amortization.
+- Standard architecture/energy cost versus biological plausibility.
+- No inference from algorithmic utility to the causal role of dendritic
+  residuals in cortex.
+
+## Reviewer objection map
+
+| Objection | Evidence that addresses it | Residual risk |
+|:--|:--|:--|
+| “This is Lansdell synthetic feedback renamed.” | `NOVELTY.md`; no-traffic method explicitly attributed; raw/matched/innovation panel isolates the new operation | Novelty remains narrow and needs clear writing |
+| “The residual only clips an oversized signal.” | Per-example norm-matched raw control; positive scaling preserves cosine; Figure mechanism panel | Artificial predictable traffic is still a controlled construction |
+| “Depth does not help this task.” | Claim says preservation; C2 reports the learned-vectorizer failure; Oral-A targets standard useful scale | Fatal to a broad scaling claim unless Oral-A passes |
+| “Local methods hide enormous extra work.” | Logical queries, MACs, peak memory, wall time, C3 confirmation, EP/native protocols | Hardware implementations are not all equally optimized |
+| “Weak baselines define the win.” | BP/FA/DFA/direct NP/FF/PEPITA/EP plus native BurstCCN and Dual Prop | Native reproductions are one seed and method-native, not equal compute |
+| “Harnett already proves the biological story.” | Oral-B preregistration and complete negative result | Biological contribution is limited without new data or a new passed prediction |
+| “The gates were selected after results.” | Git-frozen protocols, untouched confirmation seeds, failed branches retained | Early inherited pilots predate the strict boundary and must remain labeled |
+
+## Result-dependent branch
+
+### If A3/A4 pass
+
+Lead with the standard ResNet result, promote empirical support, and retain the
+MLP useful-depth failure as a diagnosis of the earlier vectorizer. The paper
+can plausibly claim a rare local-learning method that retains performance and
+alignment with standard depth under an audited cost budget. Reviewer score
+should be reassessed from the frozen confirmation only.
+
+### If A3 or A4 fails
+
+Do not soften the threshold or replace the seed panel. Keep a mechanism paper:
+innovation is load-bearing under identifiable mixed traffic and the inherited
+causal-feedback backbone preserves depth on the controlled task, but standard
+useful scaling remains unresolved. The correct response is a narrower title,
+claim, and score—not another post-hoc ResNet tuning branch.