Automaton-Supervised Neural Policy Shield / results.json
Mechanism failed
1{
2 "seed": 2516,
3 "printed_ABt_certificate_dimensionally_ill_typed": true,
4 "exact_enabled_pair_certificate": true,
5 "transition_formula_ok": true,
6 "prediction_zero_forbidden": true,
7 "prediction_first_intervention_matches_softmax": false,
8 "horizon_sweep": [
9 {
10 "H": 1,
11 "violation_rate": 0.0,
12 "completion_rate": 0.0,
13 "intervention_rate": 0.8625,
14 "first_intervention_rate": 0.021
15 },
16 {
17 "H": 2,
18 "violation_rate": 0.0,
19 "completion_rate": 1.0,
20 "intervention_rate": 0.9835,
21 "first_intervention_rate": 0.982
22 },
23 {
24 "H": 3,
25 "violation_rate": 0.0,
26 "completion_rate": 1.0,
27 "intervention_rate": 0.9835,
28 "first_intervention_rate": 0.982
29 },
30 {
31 "H": 4,
32 "violation_rate": 0.0,
33 "completion_rate": 1.0,
34 "intervention_rate": 0.9835,
35 "first_intervention_rate": 0.982
36 }
37 ],
38 "bias_sweep": [
39 {
40 "bias": -2,
41 "predicted_start_mask_probability": 0.10650697891920076,
42 "observed_first_intervention": 0.35975,
43 "shield_forbidden": 0.0,
44 "raw_forbidden": 0.5965,
45 "shield_completion": 0.87775
46 },
47 {
48 "bias": -1,
49 "predicted_start_mask_probability": 0.24472847105479764,
50 "observed_first_intervention": 0.49225,
51 "shield_forbidden": 0.0,
52 "raw_forbidden": 0.7035,
53 "shield_completion": 0.73975
54 },
55 {
56 "bias": 0,
57 "predicted_start_mask_probability": 0.4683105308334812,
58 "observed_first_intervention": 0.56475,
59 "shield_forbidden": 0.0,
60 "raw_forbidden": 0.841,
61 "shield_completion": 0.5055
62 },
63 {
64 "bias": 1,
65 "predicted_start_mask_probability": 0.7053845126982411,
66 "observed_first_intervention": 0.4315,
67 "shield_forbidden": 0.0,
68 "raw_forbidden": 0.94325,
69 "shield_completion": 0.27
70 },
71 {
72 "bias": 2,
73 "predicted_start_mask_probability": 0.8668133321973349,
74 "observed_first_intervention": 0.2315,
75 "shield_forbidden": 0.0,
76 "raw_forbidden": 0.98575,
77 "shield_completion": 0.114
78 },
79 {
80 "bias": 3,
81 "predicted_start_mask_probability": 0.9464991225528935,
82 "observed_first_intervention": 0.1015,
83 "shield_forbidden": 0.0,
84 "raw_forbidden": 0.99875,
85 "shield_completion": 0.048
86 },
87 {
88 "bias": 4,
89 "predicted_start_mask_probability": 0.9796292071670795,
90 "observed_first_intervention": 0.037,
91 "shield_forbidden": 0.0,
92 "raw_forbidden": 1.0,
93 "shield_completion": 0.01675
94 },
95 {
96 "bias": 5,
97 "predicted_start_mask_probability": 0.9924082466500291,
98 "observed_first_intervention": 0.01475,
99 "shield_forbidden": 0.0,
100 "raw_forbidden": 1.0,
101 "shield_completion": 0.00825
102 }
103 ],
104 "matrices": {
105 "A": [
106 [
107 1,
108 0,
109 0,
110 0
111 ],
112 [
113 1,
114 1,
115 0,
116 0
117 ],
118 [
119 1,
120 0,
121 1,
122 0
123 ],
124 [
125 0,
126 0,
127 1,
128 1
129 ]
130 ],
131 "B": [
132 [
133 1,
134 0,
135 0,
136 0,
137 0,
138 0,
139 0
140 ],
141 [
142 0,
143 1,
144 0,
145 1,
146 0,
147 0,
148 0
149 ],
150 [
151 0,
152 0,
153 1,
154 0,
155 1,
156 0,
157 0
158 ],
159 [
160 0,
161 0,
162 0,
163 0,
164 0,
165 1,
166 1
167 ]
168 ],
169 "C": [
170 [
171 1,
172 1,
173 1,
174 0,
175 0,
176 0,
177 0
178 ],
179 [
180 0,
181 0,
182 0,
183 1,
184 0,
185 0,
186 0
187 ],
188 [
189 0,
190 0,
191 0,
192 0,
193 1,
194 1,
195 0
196 ],
197 [
198 0,
199 0,
200 0,
201 0,
202 0,
203 0,
204 1
205 ]
206 ]
207 }
208}