Certainty-Equivalent Auxiliary Critic / results.json
Mechanism failed
1{
2 "math_check": {
3 "logmeanexp_absolute_difference": 9.928565702921333e-08,
4 "sample_scaling": [
5 {
6 "K": 2,
7 "abs_error_mean": 0.7157605508005723,
8 "bias": 0.436480942274036
9 },
10 {
11 "K": 8,
12 "abs_error_mean": 0.40372897433680116,
13 "bias": 0.1729428094129632
14 },
15 {
16 "K": 32,
17 "abs_error_mean": 0.2113885934796732,
18 "bias": 0.05154215594925562
19 },
20 {
21 "K": 128,
22 "abs_error_mean": 0.11300307392602017,
23 "bias": 0.014449644809360049
24 }
25 ]
26 },
27 "runs": [
28 {
29 "mode": "direct",
30 "seed": 11,
31 "device": "cuda",
32 "seconds": 7.432541204998415,
33 "estimator_variance": 5.50183676395477e-10,
34 "value_target_rmse": 6.896183967590332,
35 "ce_rmse": NaN,
36 "checkpoints": [
37 {
38 "step": 601,
39 "ce_abs_error": NaN,
40 "bellman_rmse": 4.506957530975342,
41 "loss": 122.40324401855469
42 },
43 {
44 "step": 1200,
45 "ce_abs_error": NaN,
46 "bellman_rmse": 6.88680362701416,
47 "loss": 425.3627624511719
48 }
49 ]
50 },
51 {
52 "mode": "aux",
53 "seed": 11,
54 "device": "cuda",
55 "seconds": 9.682978046999779,
56 "estimator_variance": 2.9493247843959125e-09,
57 "value_target_rmse": 6.245853900909424,
58 "ce_rmse": 9.322518348693848,
59 "checkpoints": [
60 {
61 "step": 601,
62 "ce_abs_error": 5.964967250823975,
63 "bellman_rmse": 1.4166280031204224,
64 "loss": 66.29203033447266
65 },
66 {
67 "step": 1200,
68 "ce_abs_error": 9.322509765625,
69 "bellman_rmse": 2.374951124191284,
70 "loss": 234.13453674316406
71 }
72 ]
73 },
74 {
75 "mode": "direct",
76 "seed": 22,
77 "device": "cuda",
78 "seconds": 6.797823614997469,
79 "estimator_variance": 5.42665135050413e-10,
80 "value_target_rmse": 6.920567035675049,
81 "ce_rmse": NaN,
82 "checkpoints": [
83 {
84 "step": 601,
85 "ce_abs_error": NaN,
86 "bellman_rmse": 4.586386203765869,
87 "loss": 122.48664855957031
88 },
89 {
90 "step": 1200,
91 "ce_abs_error": NaN,
92 "bellman_rmse": 6.939746379852295,
93 "loss": 431.5095520019531
94 }
95 ]
96 },
97 {
98 "mode": "aux",
99 "seed": 22,
100 "device": "cuda",
101 "seconds": 9.508438042998023,
102 "estimator_variance": 5.449813045288465e-09,
103 "value_target_rmse": 6.109918117523193,
104 "ce_rmse": 9.53398323059082,
105 "checkpoints": [
106 {
107 "step": 601,
108 "ce_abs_error": 5.9523162841796875,
109 "bellman_rmse": 1.4783697128295898,
110 "loss": 52.603919982910156
111 },
112 {
113 "step": 1200,
114 "ce_abs_error": 9.533950805664062,
115 "bellman_rmse": 2.669316530227661,
116 "loss": 212.74722290039062
117 }
118 ]
119 },
120 {
121 "mode": "direct",
122 "seed": 33,
123 "device": "cuda",
124 "seconds": 7.001646349999646,
125 "estimator_variance": 4.765146055518699e-10,
126 "value_target_rmse": 6.83821439743042,
127 "ce_rmse": NaN,
128 "checkpoints": [
129 {
130 "step": 601,
131 "ce_abs_error": NaN,
132 "bellman_rmse": 4.4487528800964355,
133 "loss": 119.88429260253906
134 },
135 {
136 "step": 1200,
137 "ce_abs_error": NaN,
138 "bellman_rmse": 6.826839923858643,
139 "loss": 417.4648742675781
140 }
141 ]
142 },
143 {
144 "mode": "aux",
145 "seed": 33,
146 "device": "cuda",
147 "seconds": 9.544574978997844,
148 "estimator_variance": 2.3244410840561613e-09,
149 "value_target_rmse": 6.222592830657959,
150 "ce_rmse": 9.921758651733398,
151 "checkpoints": [
152 {
153 "step": 601,
154 "ce_abs_error": 6.286478042602539,
155 "bellman_rmse": 1.7113144397735596,
156 "loss": 60.145721435546875
157 },
158 {
159 "step": 1200,
160 "ce_abs_error": 9.921757698059082,
161 "bellman_rmse": 2.9235434532165527,
162 "loss": 220.10765075683594
163 }
164 ]
165 }
166 ]
167}