Conditioned Irregular-Delay State Encoder / results.json

Failed on benchmark

Raw ⬇ ZIP
  1{
  2  "seed": 2272,
  3  "A": [
  4    [
  5      -0.08,
  6      -2.0
  7    ],
  8    [
  9      2.0,
 10      -0.08
 11    ]
 12  ],
 13  "c": [
 14    1.0,
 15    0.0
 16  ],
 17  "predictions": {
 18    "noise_linear_scaling": {
 19      "predicted_slope": 1.0,
 20      "observed_loglog_slope": 0.9992093759685947,
 21      "design_smin": 1.6318562643063428,
 22      "design_condition": 1.041333806162916
 23    },
 24    "inverse_singular_value_noise_amplification": {
 25      "expected": "positive monotonic relation",
 26      "observed_spearman": 0.942857142857143,
 27      "threshold": 0.7
 28    },
 29    "delay_design_sweep": {
 30      "predicted": "non-redundant delay increases smallest singular value",
 31      "base_delays": [
 32        0.0,
 33        0.25,
 34        0.55,
 35        0.85
 36      ],
 37      "base_smin": 1.1833368554936399,
 38      "best_added_delay": 2.7800000000000002,
 39      "best_smin": 1.6965047587741895,
 40      "improvement_ratio": 1.4336617260741675,
 41      "candidate_grid_step": 0.01
 42    }
 43  },
 44  "designs": [
 45    {
 46      "name": "uniform_short",
 47      "delays": [
 48        0.0,
 49        0.0875,
 50        0.175,
 51        0.26249999999999996,
 52        0.35
 53      ],
 54      "smin": 0.5515100675915263,
 55      "smax": 2.1997324726412715,
 56      "condition": 3.9885626789146746,
 57      "predicted_rmse": 0.05607969306915909,
 58      "observed_rmse": 0.056553992445059806
 59    },
 60    {
 61      "name": "uniform_medium",
 62      "delays": [
 63        0.0,
 64        0.325,
 65        0.65,
 66        0.9750000000000001,
 67        1.3
 68      ],
 69      "smin": 1.62059882201916,
 70      "smax": 1.7136729641730906,
 71      "condition": 1.0574319448399736,
 72      "predicted_rmse": 0.02547844557626368,
 73      "observed_rmse": 0.025183242603360783
 74    },
 75    {
 76      "name": "uniform_long",
 77      "delays": [
 78        0.0,
 79        0.65,
 80        1.3,
 81        1.9500000000000002,
 82        2.6
 83      ],
 84      "smin": 1.7013805241119362,
 85      "smax": 1.8243241210204846,
 86      "condition": 1.072261081613545,
 87      "predicted_rmse": 0.024110854910254987,
 88      "observed_rmse": 0.024081439267418937
 89    },
 90    {
 91      "name": "irregular_good",
 92      "delays": [
 93        0.0,
 94        0.22,
 95        0.61,
 96        1.07,
 97        1.56
 98      ],
 99      "smin": 1.407472686120402,
100      "smax": 1.904628352102376,
101      "condition": 1.353225800318973,
102      "predicted_rmse": 0.02650317008614197,
103      "observed_rmse": 0.0262417949607761
104    },
105    {
106      "name": "irregular_bad_clustered",
107      "delays": [
108        0.0,
109        0.02,
110        0.04,
111        0.07,
112        0.11
113      ],
114      "smin": 0.17353094123029292,
115      "smax": 2.237974845505481,
116      "condition": 12.896690524691298,
117      "predicted_rmse": 0.17339875942791294,
118      "observed_rmse": 0.1711548643355998
119    },
120    {
121      "name": "irregular_multiscale",
122      "delays": [
123        0.0,
124        0.08,
125        0.4,
126        1.3,
127        2.35
128      ],
129      "smin": 1.5271030178054346,
130      "smax": 1.8532650500502241,
131      "condition": 1.2135822065976334,
132      "predicted_rmse": 0.025455205909601947,
133      "observed_rmse": 0.02556536405917574
134    }
135  ],
136  "noise_sweep": [
137    {
138      "sigma": 0.002,
139      "observed_rmse": 0.0017021601882811978,
140      "predicted_rmse": 0.0016992065244941226,
141      "worst_case_bound": 0.0017332575096302841
142    },
143    {
144      "sigma": 0.005,
145      "observed_rmse": 0.004208765756780203,
146      "predicted_rmse": 0.004248016311235307,
147      "worst_case_bound": 0.00433314377407571
148    },
149    {
150      "sigma": 0.01,
151      "observed_rmse": 0.008489553919248948,
152      "predicted_rmse": 0.008496032622470614,
153      "worst_case_bound": 0.00866628754815142
154    },
155    {
156      "sigma": 0.02,
157      "observed_rmse": 0.01689760524536054,
158      "predicted_rmse": 0.016992065244941228,
159      "worst_case_bound": 0.01733257509630284
160    },
161    {
162      "sigma": 0.05,
163      "observed_rmse": 0.0414607642676477,
164      "predicted_rmse": 0.042480163112353066,
165      "worst_case_bound": 0.04333143774075711
166    },
167    {
168      "sigma": 0.1,
169      "observed_rmse": 0.08568598885317875,
170      "predicted_rmse": 0.08496032622470613,
171      "worst_case_bound": 0.08666287548151422
172    }
173  ],
174  "spearman": {
175    "rho_predicted_factor_vs_observed_rmse": 0.942857142857143,
176    "p_value": 0.004804664723032055
177  },
178  "poisson_baseline_comparison": [
179    {
180      "mean_gap": 0.08,
181      "baseline_fixed_uniform_error": 1.0612637805125682,
182      "conditioned_actual_delay_error": 0.06300816069821169,
183      "relative_improvement": 0.9406291236399493
184    },
185    {
186      "mean_gap": 0.15,
187      "baseline_fixed_uniform_error": 0.7987178468355736,
188      "conditioned_actual_delay_error": 0.03590487052118042,
189      "relative_improvement": 0.955046865844514
190    },
191    {
192      "mean_gap": 0.3,
193      "baseline_fixed_uniform_error": 0.6476702599826588,
194      "conditioned_actual_delay_error": 0.028854825073284508,
195      "relative_improvement": 0.9554482784587683
196    },
197    {
198      "mean_gap": 0.5,
199      "baseline_fixed_uniform_error": 0.9154694875453312,
200      "conditioned_actual_delay_error": 0.024433723655873325,
201      "relative_improvement": 0.9733101714603422
202    }
203  ],
204  "bound_check": {
205    "max_ratio": 0.9956624770862955,
206    "mean_ratio": 0.576071426023016,
207    "claim": "ratio <= 1"
208  },
209  "worked_checks": {
210    "noise_slope_within_0.10": true,
211    "spearman_above_0.7": true,
212    "bound_holds": true,
213    "delay_improves_smin": true
214  },
215  "worked": true
216}