DFA + residual-branch penalty test: seed=123, lam=0.01 ep 0: ||h_L||=8.791e+00 ||g_2||=9.848e-04 acc=0.0957 ep 1: ||h_L||=8.047e+02 ||g_2||=1.064e-05 eval_acc=0.3320 test_acc=0.3279 ep 10: ||h_L||=7.988e+03 ||g_2||=1.857e-06 eval_acc=0.3584 test_acc=0.3448 ep 20: ||h_L||=1.251e+04 ||g_2||=1.450e-06 eval_acc=0.3633 test_acc=0.3597 ep 30: ||h_L||=1.337e+04 ||g_2||=1.414e-06 eval_acc=0.3672 test_acc=0.3610 FINAL test acc: 0.3610 Compare to: DFA-vanilla (3-seed mean): 0.308 DFA-shallow (3-seed mean): 0.349 DFA-frozen (3-seed mean): 0.349 BP-trainable (3-seed mean): 0.609 Saved results/dfa_pen_short/dfa_pen_lam0.01_s123.json Saved results/dfa_pen_short/dfa_pen_lam0.01_s123.pt