diff options
| -rw-r--r-- | NOVELTY.md | 27 | ||||
| -rw-r--r-- | README.md | 16 |
2 files changed, 41 insertions, 2 deletions
@@ -83,6 +83,23 @@ accuracy, it only supplies a stable feedback substrate. The candidate novelty still requires neutral somato-dendritic subtraction to outperform raw and norm-matched raw apical activity under the same mixed-traffic intervention. +The post-MT-1 fast neutral projection does not make reciprocal KP novel. It +adds a second-timescale operation to the innovation mechanism: + +```text +e0_l = a0_l - P_l(h_l) neutral residual +q_l = Cov(e0_l,h_l) / Var(h_l) current local coupling +r_l = s_l + e0_l - E[e0_l] - q_l(h_l-E[h_l]) stabilized innovation +``` + +The coefficient fit is local and instruction-off, so it does not use a task +loss, task-nudged state, downstream weight, or reverse pass. Its defensible +novelty is not “two phases” or generic normalization; it is the explicit +operator-stability role of a per-cell somato-dendritic innovation measured from +paired neutral observations. The paper must nevertheless count and disclose +the neutral microphase. It cannot describe this variant as single-phase, and +must discuss proximity to contrastive/phase-separated local-learning methods. + ## Candidate SDIL contribution The candidate contribution is the combination @@ -130,6 +147,11 @@ gain control. `THEORY.md` states the assumptions and the executable finite-sampl vectorizer does not. This identifies feedback amortization as the current engineering bottleneck; it is not a novelty claim and the direct estimator's 68.4x work precludes presenting it as the scalable algorithm. +- On the inherited reciprocal-KP substrate, dynamic paired-neutral innovation + passes a frozen 20-epoch ResNet-20 validation gate at 83.58%, versus 82.66% + for clean KP and 10% for the failed raw/norm-matched mixed-traffic controls. + This supports a short standard-scale load-bearing result, not yet a full or + independently confirmed scaling claim. ## Claims currently forbidden @@ -137,6 +159,8 @@ gain control. `THEORY.md` states the assumptions and the executable finite-sampl - Simultaneously perturbing all hidden layers is novel. - The no-traffic scaling panel demonstrates the necessity of somato-dendritic residuals. - SDIL is the first dendritic method to scale to deep vision tasks. +- Dynamic neutral projection is a single-phase rule or has zero observation + cost merely because it adds no convolutional MAC. - Reproducing the already-published Harnett/BurstCCN qualitative signatures is an oral-level biological contribution. @@ -154,3 +178,6 @@ gain control. `THEORY.md` states the assumptions and the executable finite-sampl source data and states that full data/code are available on request. 5. For a scaling contribution, target standard architectures and a simplicity/cost regime not already occupied by BurstCCN, rather than treating ImageNet alone as novelty. +6. Complete the predeclared D3 full ResNet endpoint and an independent + multi-seed confirmation before promoting the dynamic short result into a + standard-scale claim. @@ -15,6 +15,11 @@ r_l = a_l - a_hat_l Delta W_l = eta (r_l * local postsynaptic gain) h_(l-1)^T. ``` +The standard-ResNet stability branch additionally uses a fast paired neutral +observation to project any *remaining* affine neutral residual off current +soma before plasticity. This local microphase is explicitly counted; it is not +described as single-phase or free. + `A_l` is calibrated from antithetic node perturbations without reverse-mode differentiation or weight transport. That learned-feedback mechanism is inherited from Lansdell, Prakash, and Kording (2020); it is not claimed as @@ -61,7 +66,12 @@ and scaling behavior. See `NOVELTY.md` for the exact prior-art boundary. 82.66%, and its frozen full gate reaches 91.26% versus matched BP's 91.62% with 0.9997 late feedback cosine. The subsequent frozen mixed-traffic screen nevertheless makes raw, norm-matched raw, and innovation all nonfinite in - epoch 1, so no full or confirmation panel opens. + epoch 1. A separately frozen operator-stability branch then rules out fixed + predictor margins and passes a dynamic paired-neutral controller: its D1 + 352-step trajectory is finite, and D2 reaches `83.58%` after 20 epochs versus + clean KP's `82.66%` and the failed mixed-traffic controls' `10%`, at `1.3261x` + BP MACs and zero task-loss queries. This is still one short validation seed; + the predeclared 200-epoch D3 run is in progress and test remains untouched. - Native author-code fidelity is complete. BurstCCN reaches `80.10%` at its validation-selected epoch versus published `82.97 +/- 0.21%`; Dual Prop reaches `92.46%` versus published `92.41 +/- 0.07%`. Their audited walls are @@ -83,12 +93,14 @@ somatic statistic rather than arbitrary top-down context. `ROADMAP.md` and - `BASELINES.md`: matched and native-author baseline ledger; - `ROADMAP.md`: accept/oral evidence gates and their state; - `MIXED_TRAFFIC.md`: frozen standard-ResNet raw/matched/innovation accept gate; +- `DYNAMIC_INNOVATION.md`: post-failure two-timescale stability theory and + frozen D1--D3 gates; - `ORAL_A.md`: frozen standard CIFAR ResNet funnel; - `ORAL_A_V2.md`: frozen post-failure representable-subspace funnel; - `ORAL_A_V3.md`: frozen vectorizer-space causal-calibration funnel; - `REVIEW_SCORECARD.md`: adversarial ICLR-style score trajectory; - `results/figs/`: deterministic PDF/PNG main figures, captions, and a source - hash manifest, plus the audited RRM endpoint-tracking failure supplement. + hash manifest, plus audited RRM failure and dynamic-stability supplements. The three current main figures show the local-method Pareto frontier, credit assignment versus depth, and the load-bearing innovation ablation. Every final |