Primal-Dual Active-Set Optimizer Filter / bench_report.json
Mechanism confirmed, baseline not beaten
1{
2 "bench_version": 1,
3 "track": "tabular",
4 "model": "mlp_tiny",
5 "metric_direction": "lower is better",
6 "n_seeds": 8,
7 "baseline": {
8 "best_cfg": {
9 "lr": 0.006,
10 "epochs": 12
11 },
12 "sweep": [
13 {
14 "cfg": {
15 "lr": 0.001,
16 "epochs": 12
17 },
18 "mean": 133.94248580932617
19 },
20 {
21 "cfg": {
22 "lr": 0.003,
23 "epochs": 12
24 },
25 "mean": 16.007279753684998
26 },
27 {
28 "cfg": {
29 "lr": 0.006,
30 "epochs": 12
31 },
32 "mean": 14.950186729431152
33 }
34 ],
35 "full": {
36 "mean": 14.950186729431152,
37 "std": 1.5514200913682417,
38 "per_seed": [
39 14.034903526306152,
40 11.691411018371582,
41 14.283523559570312,
42 15.042709350585938,
43 16.13408851623535,
44 17.11699104309082,
45 15.280345916748047,
46 16.017520904541016
47 ],
48 "n": 8
49 }
50 },
51 "idea": {
52 "config": {
53 "lr": 0.001,
54 "alpha": 0.2,
55 "epochs": 12
56 },
57 "mean": 216.5899887084961,
58 "std": 10.786054269508782,
59 "per_seed": [
60 230.07789611816406,
61 217.97207641601562,
62 221.4623565673828,
63 199.12059020996094,
64 223.05921936035156,
65 202.85182189941406,
66 209.7367706298828,
67 228.43917846679688
68 ],
69 "n": 8,
70 "grid": [
71 {
72 "config": {
73 "lr": 0.001,
74 "alpha": 0.2,
75 "epochs": 12
76 },
77 "mean": 216.5899887084961,
78 "per_seed": [
79 230.07789611816406,
80 217.97207641601562,
81 221.4623565673828,
82 199.12059020996094,
83 223.05921936035156,
84 202.85182189941406,
85 209.7367706298828,
86 228.43917846679688
87 ]
88 },
89 {
90 "config": {
91 "lr": 0.003,
92 "alpha": 0.2,
93 "epochs": 12
94 },
95 "mean": 234.6817684173584,
96 "per_seed": [
97 249.0762176513672,
98 236.1846923828125,
99 239.8191375732422,
100 215.83428955078125,
101 241.5047607421875,
102 220.68380737304688,
103 227.49276733398438,
104 246.8584747314453
105 ]
106 },
107 {
108 "config": {
109 "lr": 0.006,
110 "alpha": 0.2,
111 "epochs": 12
112 },
113 "mean": 232.99076080322266,
114 "per_seed": [
115 247.46762084960938,
116 234.6169891357422,
117 238.2631378173828,
118 213.9429931640625,
119 239.95175170898438,
120 219.12632751464844,
121 225.2731475830078,
122 245.28411865234375
123 ]
124 }
125 ]
126 },
127 "comparison": {
128 "delta_mean": 201.63980197906494,
129 "idea_wins": 0,
130 "n_pairs": 8,
131 "per_seed_diffs": [
132 216.0429925918579,
133 206.28066539764404,
134 207.1788330078125,
135 184.077880859375,
136 206.9251308441162,
137 185.73483085632324,
138 194.45642471313477,
139 212.42165756225586
140 ],
141 "p_value": 0.0081,
142 "mde": 10.077764526365558,
143 "mde_rel_pct": 67.40895420741685,
144 "verdict": "idea worse (significant)",
145 "system_worked": false
146 },
147 "mechanism_signature": {
148 "predicted_mean_active": 3.9270833333333335,
149 "observed_mean_active": 3.9270833333333335,
150 "mean_iterations": 1.3046875,
151 "one_iteration_fraction": 0.8307291666666666,
152 "active_set_changes": 65,
153 "max_violation": 0.8067416300048826,
154 "confirmed": true
155 },
156 "baseline_sweep": [
157 {
158 "config": {
159 "lr": 0.001,
160 "epochs": 12
161 },
162 "mean": 133.94248580932617,
163 "per_seed": [
164 143.12049865722656,
165 131.5668487548828,
166 139.70982360839844,
167 120.17390441894531,
168 137.71632385253906,
169 123.82814025878906,
170 129.80224609375,
171 145.62210083007812
172 ]
173 },
174 {
175 "config": {
176 "lr": 0.003,
177 "epochs": 12
178 },
179 "mean": 16.007279753684998,
180 "per_seed": [
181 17.079910278320312,
182 13.425646781921387,
183 15.75783920288086,
184 16.13398551940918,
185 15.73672103881836,
186 15.811994552612305,
187 16.257793426513672,
188 17.854347229003906
189 ]
190 },
191 {
192 "config": {
193 "lr": 0.006,
194 "epochs": 12
195 },
196 "mean": 14.950186729431152,
197 "per_seed": [
198 14.034903526306152,
199 11.691411018371582,
200 14.283523559570312,
201 15.042709350585938,
202 16.13408851623535,
203 17.11699104309082,
204 15.280345916748047,
205 16.017520904541016
206 ]
207 }
208 ],
209 "idea_grid": [
210 {
211 "config": {
212 "lr": 0.001,
213 "alpha": 0.2,
214 "epochs": 12
215 },
216 "mean": 216.5899887084961,
217 "per_seed": [
218 230.07789611816406,
219 217.97207641601562,
220 221.4623565673828,
221 199.12059020996094,
222 223.05921936035156,
223 202.85182189941406,
224 209.7367706298828,
225 228.43917846679688
226 ]
227 },
228 {
229 "config": {
230 "lr": 0.003,
231 "alpha": 0.2,
232 "epochs": 12
233 },
234 "mean": 234.6817684173584,
235 "per_seed": [
236 249.0762176513672,
237 236.1846923828125,
238 239.8191375732422,
239 215.83428955078125,
240 241.5047607421875,
241 220.68380737304688,
242 227.49276733398438,
243 246.8584747314453
244 ]
245 },
246 {
247 "config": {
248 "lr": 0.006,
249 "alpha": 0.2,
250 "epochs": 12
251 },
252 "mean": 232.99076080322266,
253 "per_seed": [
254 247.46762084960938,
255 234.6169891357422,
256 238.2631378173828,
257 213.9429931640625,
258 239.95175170898438,
259 219.12632751464844,
260 225.2731475830078,
261 245.28411865234375
262 ]
263 }
264 ],
265 "search_space_parity": true
266}