From a9767ed7cd9d5ee4e1e8dd001d551d6650e36b5b Mon Sep 17 00:00:00 2001 From: YurenHao0426 Date: Wed, 22 Jul 2026 00:15:17 -0500 Subject: figures: finalize audited Pareto and scaling claims --- results/figs/main_figure_captions.md | 24 ++++++++++++++++++++++++ 1 file changed, 24 insertions(+) create mode 100644 results/figs/main_figure_captions.md (limited to 'results/figs/main_figure_captions.md') diff --git a/results/figs/main_figure_captions.md b/results/figs/main_figure_captions.md new file mode 100644 index 0000000..899fadf --- /dev/null +++ b/results/figs/main_figure_captions.md @@ -0,0 +1,24 @@ +# Main figure captions + +**Figure 1 | Accuracy–cost trade-offs for local learning.** Large points and error bars show mean ± +sample standard deviation across five initialization seeds; small translucent points show the +individual seeds. Wall time was measured on a single GTX 1080. **a,** CIFAR-10 test accuracy versus +wall time for width-64 residual MLPs trained for +five epochs. The black line is the mean-based nondominated frontier among local-learning methods. +BP is a nonlocal reference and is excluded from frontier construction. Labels give hidden depth. +**b,** SDIL versus canonical equilibrium propagation (EP) on exact 784–500–10 (d1) and +784–500–500–10 (d2) forward architectures and the same first 50,000 MNIST training examples. +Within each matched architecture, every method uses the same budget: 25 epochs for d1 and 60 epochs +for d2. EP retains its published raw-pixel, hard-sigmoid energy dynamics and relaxation +schedule; SDIL retains its z-scored, tanh feedforward dynamics. Thus architecture, examples, and +epochs are matched, but preprocessing and state dynamics are intentionally method-native. EP +training particles persist across presentations as published; test particles are freshly +zero-initialized and run for the full free-phase schedule at each evaluation. + +**Figure 2 | Credit-assignment scaling in deep local-learning networks.** Mean ± sample standard +deviation across five seeds on the same CIFAR-10 width-64 residual MLP protocol. **a,** Test +accuracy versus hidden depth. **b,** Paired accuracy gap from the exact-BP model with the same +depth and initialization seed. **c,** Mean per-sample cosine between the teaching signal and exact +descent direction, averaged over the earliest third of hidden layers. Exact gradients are used only +for this diagnostic, never for local learning. “Scaling” here means preserving useful accuracy and +credit assignment as depth grows; flattened-CIFAR accuracy itself does not increase with depth. -- cgit v1.2.3