Backward-Reachability Distance Head / results.json
Mechanism failed
1{
2 "device": "cuda",
3 "math_check": {
4 "distance_table": [
5 [
6 0.0,
7 1.0,
8 2.0,
9 3.0,
10 4.0,
11 5.0,
12 6.0,
13 7.0,
14 8.0
15 ],
16 [
17 8.0,
18 7.0,
19 6.0,
20 5.0,
21 4.0,
22 3.0,
23 2.0,
24 1.0,
25 0.0
26 ]
27 ],
28 "shortest_path_assertions": [
29 true,
30 true
31 ],
32 "greedy_progress_violations": 0,
33 "ideal_progress_hinge_sum": 0.0,
34 "math_pass": true
35 },
36 "baseline_runs": [
37 {
38 "train_reward_last40": 0.675,
39 "eval_success": 0.5,
40 "eval_steps_success_only": 4.0,
41 "critic_loss_last40": 0.07858757711946965,
42 "distance_abs_error_last40": null
43 },
44 {
45 "train_reward_last40": 0.75,
46 "eval_success": 0.5,
47 "eval_steps_success_only": 4.0,
48 "critic_loss_last40": 0.06681055419612676,
49 "distance_abs_error_last40": null
50 }
51 ],
52 "idea_runs": [
53 {
54 "train_reward_last40": 0.775,
55 "eval_success": 0.5,
56 "eval_steps_success_only": 4.0,
57 "critic_loss_last40": 0.07392901862040162,
58 "distance_abs_error_last40": 2.983294904232025
59 },
60 {
61 "train_reward_last40": 0.725,
62 "eval_success": 0.5,
63 "eval_steps_success_only": 4.0,
64 "critic_loss_last40": 0.0675667256815359,
65 "distance_abs_error_last40": 2.5086710512638093
66 }
67 ],
68 "baseline": {
69 "train_reward_last40": 0.7125,
70 "eval_success": 0.5,
71 "eval_steps_success_only": 4.0,
72 "critic_loss_last40": 0.07269906565779821
73 },
74 "idea": {
75 "train_reward_last40": 0.75,
76 "eval_success": 0.5,
77 "eval_steps_success_only": 4.0,
78 "critic_loss_last40": 0.07074787215096875,
79 "distance_abs_error_last40": 2.745982977747917
80 }
81}