Differentiable Persistence Landscape Layer / bench_report.json

Failed on benchmark

Raw ⬇ ZIP
  1{
  2  "bench_version": 1,
  3  "track": "persistence_diagrams",
  4  "model": "mlp_tiny",
  5  "metric_direction": "lower is better",
  6  "n_seeds": 8,
  7  "baseline": {
  8    "best_cfg": {
  9      "lr": 0.003
 10    },
 11    "sweep": [
 12      {
 13        "cfg": {
 14          "lr": 0.001
 15        },
 16        "mean": 0.03624999849125743
 17      },
 18      {
 19        "cfg": {
 20          "lr": 0.003
 21        },
 22        "mean": 0.0024999999441206455
 23      },
 24      {
 25        "cfg": {
 26          "lr": 0.01
 27        },
 28        "mean": 0.0024999999441206455
 29      }
 30    ],
 31    "full": {
 32      "mean": 0.0012499999720603228,
 33      "std": 0.002165063461068156,
 34      "per_seed": [
 35        0.0,
 36        0.004999999888241291,
 37        0.0,
 38        0.004999999888241291,
 39        0.0,
 40        0.0,
 41        0.0,
 42        0.0
 43      ],
 44      "n": 8
 45    }
 46  },
 47  "idea": {
 48    "mean": 0.0037499999161809683,
 49    "std": 0.0033071890649093005,
 50    "per_seed": [
 51      0.0,
 52      0.004999999888241291,
 53      0.004999999888241291,
 54      0.009999999776482582,
 55      0.0,
 56      0.004999999888241291,
 57      0.0,
 58      0.004999999888241291
 59    ],
 60    "n": 8
 61  },
 62  "comparison": {
 63    "delta_mean": 0.0024999999441206455,
 64    "idea_wins": 0,
 65    "n_pairs": 8,
 66    "per_seed_diffs": [
 67      0.0,
 68      0.0,
 69      0.004999999888241291,
 70      0.004999999888241291,
 71      0.0,
 72      0.004999999888241291,
 73      0.0,
 74      0.004999999888241291
 75    ],
 76    "p_value": 0.11955,
 77    "mde": 0.0022347148967172843,
 78    "mde_rel_pct": 178.77719573336447,
 79    "verdict": "no significant win",
 80    "system_worked": false
 81  },
 82  "mechanism_signature": {
 83    "custom_track": {
 84      "name": "persistence_diagrams",
 85      "file": "persistence_track.py",
 86      "domain": "topological_representation"
 87    },
 88    "idea_config": {
 89      "lr": 0.001
 90    },
 91    "idea_sweep": [
 92      {
 93        "cfg": {
 94          "lr": 0.001
 95        },
 96        "result": {
 97          "mean": 0.0037499999161809683,
 98          "std": 0.0033071890649093005,
 99          "per_seed": [
100            0.0,
101            0.004999999888241291,
102            0.004999999888241291,
103            0.009999999776482582,
104            0.0,
105            0.004999999888241291,
106            0.0,
107            0.004999999888241291
108          ],
109          "n": 8
110        }
111      },
112      {
113        "cfg": {
114          "lr": 0.003
115        },
116        "result": {
117          "mean": 0.0037499999161809683,
118          "std": 0.0033071890649093005,
119          "per_seed": [
120            0.0,
121            0.004999999888241291,
122            0.0,
123            0.009999999776482582,
124            0.004999999888241291,
125            0.004999999888241291,
126            0.0,
127            0.004999999888241291
128          ],
129          "n": 8
130        }
131      },
132      {
133        "cfg": {
134          "lr": 0.01
135        },
136        "result": {
137          "mean": 0.0037499999161809683,
138          "std": 0.0033071890649093005,
139          "per_seed": [
140            0.004999999888241291,
141            0.004999999888241291,
142            0.0,
143            0.009999999776482582,
144            0.004999999888241291,
145            0.004999999888241291,
146            0.0,
147            0.0
148          ],
149          "n": 8
150        }
151      }
152    ],
153    "mechanism_signature": {
154      "prediction": "landscape perturbation does not exceed matched tent perturbation",
155      "predicted_ratio_bound": 1.0,
156      "observed_ratio": 1.3417700519969626,
157      "observed_landscape_rms": 0.0025743090081959963,
158      "matched_tent_rms": 0.0019185917917639017,
159      "confirmed": false
160    }
161  }
162}