paper v2.31: matched 30-epoch BP/DFA controls (was unsourced 0.609/0.308)

The §5 ¶3 BP-no-penalty value of 0.609 ± 0.004 and DFA-no-penalty value
of 0.308 ± 0.014 turned out to be unsourced — they were carried over
from a hardcoded comment in experiments/bp_with_penalty_control.py
("BP-trainable (3-seed mean): 0.609") that nobody had actually measured
with a matched 30-epoch run.

Ran the missing matched controls under the same recipe as BP+pen
(lam=0, 30 epochs, AdamW 1e-3, wd 0.01, cosine schedule, batch 128,
3 seeds 42/123/456):

  BP no-pen 30ep: per-seed 0.5851, 0.5845, 0.5863  →  0.585 ± 0.001
                  (paper said 0.609 ± 0.004, off by 0.024)
  DFA no-pen 30ep: per-seed 0.3070, 0.2985, 0.2966 →  0.301 ± 0.005
                  (paper said 0.308 ± 0.014)

Also re-grounded DFA+penalty 30ep using the dfa_pen_short 3-seed run
(0.3593, 0.3610, 0.3604 → 0.360 ± 0.001), which is what the deep-cosine
+0.155 figure was computed on. The paper had 0.363 ± 0.001 — that came
from the 100-epoch run, not the 30-epoch run, so it was an apples-to-
oranges comparison with BP+pen 30-ep.

Paper changes (§5 ¶3):
  BP penalty cost:  -8 pp  →  -5.5 pp
  DFA pen rescue:   +5.5 → +5.9 pp
  DFA+pen margin vs frozen: +1.4 → +1.1 pp
  BP-to-DFA gap:     17 → 17.0 pp (unchanged)
  BP-to-SB gap:      7.7 → 7.7 pp (unchanged)
  BP-to-DFA gap is still the lower-bound credit-quality cost claim;
  17 pp gap is unchanged in magnitude.

Also updated:
- §5 ¶1 prose: 0.363 → 0.360, 0.308 → 0.301
- §4 ¶4 prose: DFA+pen 0.363 → 0.360
- Appendix J Table 9 caption: 0.363 → 0.360, +9.0 → +9.3 pp gap to SB
- Appendix L paragraph: +5.5 → +5.9 pp DFA penalty rescue
- Figure 3 panel C bar values + title pen-cost annotation
- New results/matched_30ep_control_summary.json as auditable record

Page layout preserved: 9 main pages + refs p10, 18 total, 0 overfull.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

1 files changed, 71 insertions, 0 deletions

diff --git a/results/matched_30ep_control_summary.json b/results/matched_30ep_control_summary.json
new file mode 100644
index 0000000..23acca3
--- /dev/null
+++ b/results/matched_30ep_control_summary.json
@@ -0,0 +1,71 @@
+{
+  "matched_30ep_control": {
+    "config": {
+      "arch": "4-block d=256 pre-LN ResMLP",
+      "dataset": "CIFAR-10",
+      "epochs": 30,
+      "batch": 128,
+      "optimizer": "AdamW lr=1e-3 wd=0.01",
+      "schedule": "cosine",
+      "penalty": "lam*mean(||f_l(h_l)||^2) on residual branches"
+    },
+    "bp_no_pen_3seed": {
+      "per_seed": {
+        "42": 0.5851,
+        "123": 0.5845,
+        "456": 0.5863
+      },
+      "mean": 0.5853,
+      "std_ddof0": 0.00075,
+      "std_ddof1": 0.00092,
+      "source_files": [
+        "results/bp_no_penalty_30ep/bp_pen_lam0.0_s{42,123,456}.json"
+      ]
+    },
+    "bp_with_pen_s42": {
+      "acc": 0.5303,
+      "lam": 0.01,
+      "source": "results/bp_with_penalty/bp_pen_lam0.01_s42.json",
+      "note": "single-seed (s42 only); v2.30 polish identified the matched 30-ep no-pen 3-seed control was missing"
+    },
+    "dfa_no_pen_3seed": {
+      "per_seed": {
+        "42": 0.307,
+        "123": 0.2985,
+        "456": 0.2966
+      },
+      "mean": 0.3007,
+      "std_ddof0": 0.0045,
+      "std_ddof1": 0.0055,
+      "source_files": [
+        "results/dfa_no_penalty_30ep/results_cifar10.json"
+      ]
+    },
+    "dfa_with_pen_3seed": {
+      "per_seed": {
+        "42": 0.3593,
+        "123": 0.361,
+        "456": 0.3604
+      },
+      "mean": 0.3602,
+      "std_ddof0": 0.0007,
+      "std_ddof1": 0.0009,
+      "source_files": [
+        "results/dfa_pen_short/dfa_pen_lam0.01_s{42,123,456}.json"
+      ],
+      "lam": 0.01
+    },
+    "penalty_costs": {
+      "bp_pen_cost_pp": 5.5,
+      "dfa_pen_rescue_pp": 5.9
+    },
+    "margins_vs_frozen_0.349": {
+      "bp_no_pen_pp": 23.6,
+      "bp_with_pen_pp": 18.1,
+      "dfa_no_pen_pp": -4.8,
+      "dfa_with_pen_pp": 1.1,
+      "sb_with_pen_pp": 10.4,
+      "cb_with_pen_pp": 1.1
+    }
+  }
+}
\ No newline at end of file