anhtld commited on
Commit
6523010
·
verified ·
1 Parent(s): e3d8534

ctt train-calibrated results 2026-07-03T16:40:14Z: runs/ctt_base_context_obs_learned_dominance_success_train_to_test

Browse files
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/command.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ python scripts/eval_learned_dominance_selector.py --calibration-input runs/ctt_base_context_obs_train_cal_rollout_comparison/combined_measured_candidates.json --calibration-target-index data/cil_charts_rgb_refs/train/index.json --eval-input runs/ctt_base_context_obs_test_rollout_comparison/combined_measured_candidates.json --eval-target-index data/cil_charts_rgb_refs/test/index.json --checkpoint-template runs/ctt_residual_base_context_obs_seed{seed}/model.pt --out-dir runs/ctt_base_context_obs_learned_dominance_success_train_to_test --feature-set basic --target success --k 8 --bootstrap-samples 1000
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/config.yaml ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ bootstrap_samples: 1000
2
+ calibration_input: runs/ctt_base_context_obs_train_cal_rollout_comparison/combined_measured_candidates.json
3
+ calibration_target_index: data/cil_charts_rgb_refs/train/index.json
4
+ checkpoint_template: runs/ctt_residual_base_context_obs_seed{seed}/model.pt
5
+ eval_input: runs/ctt_base_context_obs_test_rollout_comparison/combined_measured_candidates.json
6
+ eval_target_index: data/cil_charts_rgb_refs/test/index.json
7
+ feature_set: basic
8
+ k: 8
9
+ out_dir: runs/ctt_base_context_obs_learned_dominance_success_train_to_test
10
+ ridge_lambdas: 0,0.01,0.1,1,10,100
11
+ target: success
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/data_hash.txt ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_input": "3ba12f974e07c55307cee1da7eb33e96150557cde08248561b2e69cf0fca1f86",
3
+ "calibration_target_index": "917d372cd8ba921f6157abf6914ba32116a6c95857f04a39bed16b4563b01eb9",
4
+ "eval_input": "855f6d9ced75c3e0ced1d598ad36d72141eea98936dcee39b7bb1274242a9bb1",
5
+ "eval_target_index": "5c726c14d088a48c6a37c43257ea455f5d3ae37708c113c666118441463cff8b"
6
+ }
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/eval.log ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ evaluated learned fallback rule on held-out measured rollout rows
2
+ eval_input=runs/ctt_base_context_obs_test_rollout_comparison/combined_measured_candidates.json
3
+ num_eval_rows=144
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/git_hash.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ 14e81292e2004925f81e0bb4810f6f144bfd87b7
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/metrics.json ADDED
The diff for this file is too large to render. See raw diff
 
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/metrics_by_seed.json ADDED
@@ -0,0 +1,2762 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "1073161778/train0": {
3
+ "base_success": 1.0,
4
+ "base_utility": 2.0,
5
+ "coverage": 1.0,
6
+ "execute_generated": 1.0,
7
+ "fallback_rate": 0.0,
8
+ "hidden_chart_oracle_success": 1.0,
9
+ "hidden_chart_oracle_utility": 2.0,
10
+ "outcome_ptr": 0.0,
11
+ "predicted_margin": 0.32853129065840025,
12
+ "proposal_oracle_success": 1.0,
13
+ "proposal_oracle_utility": 2.0,
14
+ "selected_candidate_index": 4.0,
15
+ "selected_success": 0.0,
16
+ "selected_success_gain_over_base": -1.0,
17
+ "selected_utility": 0.6231040358543396,
18
+ "selected_utility_gain_over_base": -1.3768959641456604,
19
+ "selector_regret": 1.3768959641456604,
20
+ "success_selector_gap": 1.0,
21
+ "success_support_gap": 0.0,
22
+ "support_gap": 0.0,
23
+ "tau": 0.288910690801741
24
+ },
25
+ "1073161778/train1": {
26
+ "base_success": 1.0,
27
+ "base_utility": 2.0,
28
+ "coverage": 1.0,
29
+ "execute_generated": 1.0,
30
+ "fallback_rate": 0.0,
31
+ "hidden_chart_oracle_success": 1.0,
32
+ "hidden_chart_oracle_utility": 2.0,
33
+ "outcome_ptr": 0.0,
34
+ "predicted_margin": 0.29150329162364197,
35
+ "proposal_oracle_success": 0.0,
36
+ "proposal_oracle_utility": 0.6944456100463867,
37
+ "selected_candidate_index": 2.0,
38
+ "selected_success": 0.0,
39
+ "selected_success_gain_over_base": -1.0,
40
+ "selected_utility": 0.5649657845497131,
41
+ "selected_utility_gain_over_base": -1.4350342154502869,
42
+ "selector_regret": 0.12947982549667358,
43
+ "success_selector_gap": 0.0,
44
+ "success_support_gap": 1.0,
45
+ "support_gap": 1.3055543899536133,
46
+ "tau": 0.288910690801741
47
+ },
48
+ "1073161778/train2": {
49
+ "base_success": 1.0,
50
+ "base_utility": 2.0,
51
+ "coverage": 1.0,
52
+ "execute_generated": 1.0,
53
+ "fallback_rate": 0.0,
54
+ "hidden_chart_oracle_success": 1.0,
55
+ "hidden_chart_oracle_utility": 2.0,
56
+ "outcome_ptr": 0.0,
57
+ "predicted_margin": 0.31124061123557867,
58
+ "proposal_oracle_success": 1.0,
59
+ "proposal_oracle_utility": 2.0,
60
+ "selected_candidate_index": 3.0,
61
+ "selected_success": 1.0,
62
+ "selected_success_gain_over_base": 0.0,
63
+ "selected_utility": 2.0,
64
+ "selected_utility_gain_over_base": 0.0,
65
+ "selector_regret": 0.0,
66
+ "success_selector_gap": 0.0,
67
+ "success_support_gap": 0.0,
68
+ "support_gap": 0.0,
69
+ "tau": 0.288910690801741
70
+ },
71
+ "1113371261/train0": {
72
+ "base_success": 0.0,
73
+ "base_utility": 0.15842342376708984,
74
+ "coverage": 1.0,
75
+ "execute_generated": 1.0,
76
+ "fallback_rate": 0.0,
77
+ "hidden_chart_oracle_success": 1.0,
78
+ "hidden_chart_oracle_utility": 2.0,
79
+ "outcome_ptr": 0.0,
80
+ "predicted_margin": 0.29841976589232927,
81
+ "proposal_oracle_success": 0.0,
82
+ "proposal_oracle_utility": 0.13477009534835815,
83
+ "selected_candidate_index": 3.0,
84
+ "selected_success": 0.0,
85
+ "selected_success_gain_over_base": 0.0,
86
+ "selected_utility": 0.09366293251514435,
87
+ "selected_utility_gain_over_base": -0.0647604912519455,
88
+ "selector_regret": 0.041107162833213806,
89
+ "success_selector_gap": 0.0,
90
+ "success_support_gap": 1.0,
91
+ "support_gap": 1.8652299046516418,
92
+ "tau": 0.288910690801741
93
+ },
94
+ "1113371261/train1": {
95
+ "base_success": 0.0,
96
+ "base_utility": 0.15842342376708984,
97
+ "coverage": 0.0,
98
+ "execute_generated": 0.0,
99
+ "fallback_rate": 1.0,
100
+ "hidden_chart_oracle_success": 1.0,
101
+ "hidden_chart_oracle_utility": 2.0,
102
+ "outcome_ptr": 1.0,
103
+ "predicted_margin": 0.25845846730035027,
104
+ "proposal_oracle_success": 0.0,
105
+ "proposal_oracle_utility": 0.1832391619682312,
106
+ "selected_candidate_index": 2.0,
107
+ "selected_success": 0.0,
108
+ "selected_success_gain_over_base": 0.0,
109
+ "selected_utility": 0.15842342376708984,
110
+ "selected_utility_gain_over_base": 0.0,
111
+ "selector_regret": 0.024815738201141357,
112
+ "success_selector_gap": 0.0,
113
+ "success_support_gap": 1.0,
114
+ "support_gap": 1.8167608380317688,
115
+ "tau": 0.288910690801741
116
+ },
117
+ "1113371261/train2": {
118
+ "base_success": 0.0,
119
+ "base_utility": 0.15842342376708984,
120
+ "coverage": 0.0,
121
+ "execute_generated": 0.0,
122
+ "fallback_rate": 1.0,
123
+ "hidden_chart_oracle_success": 1.0,
124
+ "hidden_chart_oracle_utility": 2.0,
125
+ "outcome_ptr": 1.0,
126
+ "predicted_margin": 0.2586284691418689,
127
+ "proposal_oracle_success": 0.0,
128
+ "proposal_oracle_utility": 0.2044394612312317,
129
+ "selected_candidate_index": 4.0,
130
+ "selected_success": 0.0,
131
+ "selected_success_gain_over_base": 0.0,
132
+ "selected_utility": 0.15842342376708984,
133
+ "selected_utility_gain_over_base": 0.0,
134
+ "selector_regret": 0.046016037464141846,
135
+ "success_selector_gap": 0.0,
136
+ "success_support_gap": 1.0,
137
+ "support_gap": 1.7955605387687683,
138
+ "tau": 0.288910690801741
139
+ },
140
+ "1151410346/train0": {
141
+ "base_success": 0.0,
142
+ "base_utility": 0.2260012924671173,
143
+ "coverage": 1.0,
144
+ "execute_generated": 1.0,
145
+ "fallback_rate": 0.0,
146
+ "hidden_chart_oracle_success": 1.0,
147
+ "hidden_chart_oracle_utility": 2.0,
148
+ "outcome_ptr": 1.0,
149
+ "predicted_margin": 0.3978107216823969,
150
+ "proposal_oracle_success": 0.0,
151
+ "proposal_oracle_utility": 0.8404104709625244,
152
+ "selected_candidate_index": 3.0,
153
+ "selected_success": 0.0,
154
+ "selected_success_gain_over_base": 0.0,
155
+ "selected_utility": 0.2033921778202057,
156
+ "selected_utility_gain_over_base": -0.02260911464691162,
157
+ "selector_regret": 0.6370182931423187,
158
+ "success_selector_gap": 0.0,
159
+ "success_support_gap": 1.0,
160
+ "support_gap": 1.1595895290374756,
161
+ "tau": 0.288910690801741
162
+ },
163
+ "1151410346/train1": {
164
+ "base_success": 0.0,
165
+ "base_utility": 0.2260012924671173,
166
+ "coverage": 1.0,
167
+ "execute_generated": 1.0,
168
+ "fallback_rate": 0.0,
169
+ "hidden_chart_oracle_success": 1.0,
170
+ "hidden_chart_oracle_utility": 2.0,
171
+ "outcome_ptr": 1.0,
172
+ "predicted_margin": 0.3400535300305296,
173
+ "proposal_oracle_success": 1.0,
174
+ "proposal_oracle_utility": 2.0,
175
+ "selected_candidate_index": 1.0,
176
+ "selected_success": 0.0,
177
+ "selected_success_gain_over_base": 0.0,
178
+ "selected_utility": 0.20698711276054382,
179
+ "selected_utility_gain_over_base": -0.019014179706573486,
180
+ "selector_regret": 1.7930128872394562,
181
+ "success_selector_gap": 1.0,
182
+ "success_support_gap": 0.0,
183
+ "support_gap": 0.0,
184
+ "tau": 0.288910690801741
185
+ },
186
+ "1151410346/train2": {
187
+ "base_success": 0.0,
188
+ "base_utility": 0.2260012924671173,
189
+ "coverage": 1.0,
190
+ "execute_generated": 1.0,
191
+ "fallback_rate": 0.0,
192
+ "hidden_chart_oracle_success": 1.0,
193
+ "hidden_chart_oracle_utility": 2.0,
194
+ "outcome_ptr": 1.0,
195
+ "predicted_margin": 0.36366471241281534,
196
+ "proposal_oracle_success": 0.0,
197
+ "proposal_oracle_utility": 0.5435987114906311,
198
+ "selected_candidate_index": 2.0,
199
+ "selected_success": 0.0,
200
+ "selected_success_gain_over_base": 0.0,
201
+ "selected_utility": 0.16552376747131348,
202
+ "selected_utility_gain_over_base": -0.06047752499580383,
203
+ "selector_regret": 0.3780749440193176,
204
+ "success_selector_gap": 0.0,
205
+ "success_support_gap": 1.0,
206
+ "support_gap": 1.456401288509369,
207
+ "tau": 0.288910690801741
208
+ },
209
+ "1261930456/train0": {
210
+ "base_success": 1.0,
211
+ "base_utility": 2.0,
212
+ "coverage": 1.0,
213
+ "execute_generated": 1.0,
214
+ "fallback_rate": 0.0,
215
+ "hidden_chart_oracle_success": 1.0,
216
+ "hidden_chart_oracle_utility": 2.0,
217
+ "outcome_ptr": 0.0,
218
+ "predicted_margin": 0.33197358360939205,
219
+ "proposal_oracle_success": 1.0,
220
+ "proposal_oracle_utility": 2.0,
221
+ "selected_candidate_index": 2.0,
222
+ "selected_success": 1.0,
223
+ "selected_success_gain_over_base": 0.0,
224
+ "selected_utility": 2.0,
225
+ "selected_utility_gain_over_base": 0.0,
226
+ "selector_regret": 0.0,
227
+ "success_selector_gap": 0.0,
228
+ "success_support_gap": 0.0,
229
+ "support_gap": 0.0,
230
+ "tau": 0.288910690801741
231
+ },
232
+ "1261930456/train1": {
233
+ "base_success": 1.0,
234
+ "base_utility": 2.0,
235
+ "coverage": 1.0,
236
+ "execute_generated": 1.0,
237
+ "fallback_rate": 0.0,
238
+ "hidden_chart_oracle_success": 1.0,
239
+ "hidden_chart_oracle_utility": 2.0,
240
+ "outcome_ptr": 0.0,
241
+ "predicted_margin": 0.32529166606353305,
242
+ "proposal_oracle_success": 1.0,
243
+ "proposal_oracle_utility": 2.0,
244
+ "selected_candidate_index": 4.0,
245
+ "selected_success": 0.0,
246
+ "selected_success_gain_over_base": -1.0,
247
+ "selected_utility": 0.05607958137989044,
248
+ "selected_utility_gain_over_base": -1.9439204186201096,
249
+ "selector_regret": 1.9439204186201096,
250
+ "success_selector_gap": 1.0,
251
+ "success_support_gap": 0.0,
252
+ "support_gap": 0.0,
253
+ "tau": 0.288910690801741
254
+ },
255
+ "1261930456/train2": {
256
+ "base_success": 1.0,
257
+ "base_utility": 2.0,
258
+ "coverage": 1.0,
259
+ "execute_generated": 1.0,
260
+ "fallback_rate": 0.0,
261
+ "hidden_chart_oracle_success": 1.0,
262
+ "hidden_chart_oracle_utility": 2.0,
263
+ "outcome_ptr": 0.0,
264
+ "predicted_margin": 0.32289934935450115,
265
+ "proposal_oracle_success": 1.0,
266
+ "proposal_oracle_utility": 2.0,
267
+ "selected_candidate_index": 2.0,
268
+ "selected_success": 1.0,
269
+ "selected_success_gain_over_base": 0.0,
270
+ "selected_utility": 2.0,
271
+ "selected_utility_gain_over_base": 0.0,
272
+ "selector_regret": 0.0,
273
+ "success_selector_gap": 0.0,
274
+ "success_support_gap": 0.0,
275
+ "support_gap": 0.0,
276
+ "tau": 0.288910690801741
277
+ },
278
+ "1324133122/train0": {
279
+ "base_success": 0.0,
280
+ "base_utility": 0.09220491349697113,
281
+ "coverage": 0.0,
282
+ "execute_generated": 0.0,
283
+ "fallback_rate": 1.0,
284
+ "hidden_chart_oracle_success": 1.0,
285
+ "hidden_chart_oracle_utility": 2.0,
286
+ "outcome_ptr": 0.0,
287
+ "predicted_margin": 0.2507616350624362,
288
+ "proposal_oracle_success": 0.0,
289
+ "proposal_oracle_utility": 0.07098388671875,
290
+ "selected_candidate_index": 5.0,
291
+ "selected_success": 0.0,
292
+ "selected_success_gain_over_base": 0.0,
293
+ "selected_utility": 0.09220491349697113,
294
+ "selected_utility_gain_over_base": 0.0,
295
+ "selector_regret": 0.0,
296
+ "success_selector_gap": 0.0,
297
+ "success_support_gap": 1.0,
298
+ "support_gap": 1.92901611328125,
299
+ "tau": 0.288910690801741
300
+ },
301
+ "1324133122/train1": {
302
+ "base_success": 0.0,
303
+ "base_utility": 0.09220491349697113,
304
+ "coverage": 1.0,
305
+ "execute_generated": 1.0,
306
+ "fallback_rate": 0.0,
307
+ "hidden_chart_oracle_success": 1.0,
308
+ "hidden_chart_oracle_utility": 2.0,
309
+ "outcome_ptr": 1.0,
310
+ "predicted_margin": 0.3184436611118522,
311
+ "proposal_oracle_success": 0.0,
312
+ "proposal_oracle_utility": 0.14303791522979736,
313
+ "selected_candidate_index": 6.0,
314
+ "selected_success": 0.0,
315
+ "selected_success_gain_over_base": 0.0,
316
+ "selected_utility": 0.08513790369033813,
317
+ "selected_utility_gain_over_base": -0.007067009806632996,
318
+ "selector_regret": 0.05790001153945923,
319
+ "success_selector_gap": 0.0,
320
+ "success_support_gap": 1.0,
321
+ "support_gap": 1.8569620847702026,
322
+ "tau": 0.288910690801741
323
+ },
324
+ "1324133122/train2": {
325
+ "base_success": 0.0,
326
+ "base_utility": 0.09220491349697113,
327
+ "coverage": 0.0,
328
+ "execute_generated": 0.0,
329
+ "fallback_rate": 1.0,
330
+ "hidden_chart_oracle_success": 1.0,
331
+ "hidden_chart_oracle_utility": 2.0,
332
+ "outcome_ptr": 1.0,
333
+ "predicted_margin": 0.2557727270247497,
334
+ "proposal_oracle_success": 0.0,
335
+ "proposal_oracle_utility": 0.18079039454460144,
336
+ "selected_candidate_index": 2.0,
337
+ "selected_success": 0.0,
338
+ "selected_success_gain_over_base": 0.0,
339
+ "selected_utility": 0.09220491349697113,
340
+ "selected_utility_gain_over_base": 0.0,
341
+ "selector_regret": 0.08858548104763031,
342
+ "success_selector_gap": 0.0,
343
+ "success_support_gap": 1.0,
344
+ "support_gap": 1.8192096054553986,
345
+ "tau": 0.288910690801741
346
+ },
347
+ "1324642411/train0": {
348
+ "base_success": 0.0,
349
+ "base_utility": 0.10647052526473999,
350
+ "coverage": 1.0,
351
+ "execute_generated": 1.0,
352
+ "fallback_rate": 0.0,
353
+ "hidden_chart_oracle_success": 1.0,
354
+ "hidden_chart_oracle_utility": 2.0,
355
+ "outcome_ptr": 1.0,
356
+ "predicted_margin": 0.32824135621261386,
357
+ "proposal_oracle_success": 0.0,
358
+ "proposal_oracle_utility": 0.16031095385551453,
359
+ "selected_candidate_index": 2.0,
360
+ "selected_success": 0.0,
361
+ "selected_success_gain_over_base": 0.0,
362
+ "selected_utility": 0.0764332264661789,
363
+ "selected_utility_gain_over_base": -0.030037298798561096,
364
+ "selector_regret": 0.08387772738933563,
365
+ "success_selector_gap": 0.0,
366
+ "success_support_gap": 1.0,
367
+ "support_gap": 1.8396890461444855,
368
+ "tau": 0.288910690801741
369
+ },
370
+ "1324642411/train1": {
371
+ "base_success": 0.0,
372
+ "base_utility": 0.10647052526473999,
373
+ "coverage": 1.0,
374
+ "execute_generated": 1.0,
375
+ "fallback_rate": 0.0,
376
+ "hidden_chart_oracle_success": 1.0,
377
+ "hidden_chart_oracle_utility": 2.0,
378
+ "outcome_ptr": 1.0,
379
+ "predicted_margin": 0.35265968189747576,
380
+ "proposal_oracle_success": 0.0,
381
+ "proposal_oracle_utility": 0.16803905367851257,
382
+ "selected_candidate_index": 1.0,
383
+ "selected_success": 0.0,
384
+ "selected_success_gain_over_base": 0.0,
385
+ "selected_utility": 0.11297215521335602,
386
+ "selected_utility_gain_over_base": 0.006501629948616028,
387
+ "selector_regret": 0.055066898465156555,
388
+ "success_selector_gap": 0.0,
389
+ "success_support_gap": 1.0,
390
+ "support_gap": 1.8319609463214874,
391
+ "tau": 0.288910690801741
392
+ },
393
+ "1324642411/train2": {
394
+ "base_success": 0.0,
395
+ "base_utility": 0.10647052526473999,
396
+ "coverage": 0.0,
397
+ "execute_generated": 0.0,
398
+ "fallback_rate": 1.0,
399
+ "hidden_chart_oracle_success": 1.0,
400
+ "hidden_chart_oracle_utility": 2.0,
401
+ "outcome_ptr": 1.0,
402
+ "predicted_margin": 0.22752747502297607,
403
+ "proposal_oracle_success": 1.0,
404
+ "proposal_oracle_utility": 2.0,
405
+ "selected_candidate_index": 3.0,
406
+ "selected_success": 0.0,
407
+ "selected_success_gain_over_base": 0.0,
408
+ "selected_utility": 0.10647052526473999,
409
+ "selected_utility_gain_over_base": 0.0,
410
+ "selector_regret": 1.89352947473526,
411
+ "success_selector_gap": 1.0,
412
+ "success_support_gap": 0.0,
413
+ "support_gap": 0.0,
414
+ "tau": 0.288910690801741
415
+ },
416
+ "1375492786/train0": {
417
+ "base_success": 1.0,
418
+ "base_utility": 2.0,
419
+ "coverage": 0.0,
420
+ "execute_generated": 0.0,
421
+ "fallback_rate": 1.0,
422
+ "hidden_chart_oracle_success": 1.0,
423
+ "hidden_chart_oracle_utility": 2.0,
424
+ "outcome_ptr": 0.0,
425
+ "predicted_margin": 0.21305097215459415,
426
+ "proposal_oracle_success": 1.0,
427
+ "proposal_oracle_utility": 2.0,
428
+ "selected_candidate_index": 2.0,
429
+ "selected_success": 1.0,
430
+ "selected_success_gain_over_base": 0.0,
431
+ "selected_utility": 2.0,
432
+ "selected_utility_gain_over_base": 0.0,
433
+ "selector_regret": 0.0,
434
+ "success_selector_gap": 0.0,
435
+ "success_support_gap": 0.0,
436
+ "support_gap": 0.0,
437
+ "tau": 0.288910690801741
438
+ },
439
+ "1375492786/train1": {
440
+ "base_success": 1.0,
441
+ "base_utility": 2.0,
442
+ "coverage": 0.0,
443
+ "execute_generated": 0.0,
444
+ "fallback_rate": 1.0,
445
+ "hidden_chart_oracle_success": 1.0,
446
+ "hidden_chart_oracle_utility": 2.0,
447
+ "outcome_ptr": 0.0,
448
+ "predicted_margin": 0.1579738139117352,
449
+ "proposal_oracle_success": 1.0,
450
+ "proposal_oracle_utility": 2.0,
451
+ "selected_candidate_index": 5.0,
452
+ "selected_success": 1.0,
453
+ "selected_success_gain_over_base": 0.0,
454
+ "selected_utility": 2.0,
455
+ "selected_utility_gain_over_base": 0.0,
456
+ "selector_regret": 0.0,
457
+ "success_selector_gap": 0.0,
458
+ "success_support_gap": 0.0,
459
+ "support_gap": 0.0,
460
+ "tau": 0.288910690801741
461
+ },
462
+ "1375492786/train2": {
463
+ "base_success": 1.0,
464
+ "base_utility": 2.0,
465
+ "coverage": 0.0,
466
+ "execute_generated": 0.0,
467
+ "fallback_rate": 1.0,
468
+ "hidden_chart_oracle_success": 1.0,
469
+ "hidden_chart_oracle_utility": 2.0,
470
+ "outcome_ptr": 0.0,
471
+ "predicted_margin": 0.15362020521362918,
472
+ "proposal_oracle_success": 0.0,
473
+ "proposal_oracle_utility": 0.10372558981180191,
474
+ "selected_candidate_index": 5.0,
475
+ "selected_success": 1.0,
476
+ "selected_success_gain_over_base": 0.0,
477
+ "selected_utility": 2.0,
478
+ "selected_utility_gain_over_base": 0.0,
479
+ "selector_regret": 0.0,
480
+ "success_selector_gap": 0.0,
481
+ "success_support_gap": 1.0,
482
+ "support_gap": 1.896274410188198,
483
+ "tau": 0.288910690801741
484
+ },
485
+ "141984615/train0": {
486
+ "base_success": 0.0,
487
+ "base_utility": 0.10283424705266953,
488
+ "coverage": 1.0,
489
+ "execute_generated": 1.0,
490
+ "fallback_rate": 0.0,
491
+ "hidden_chart_oracle_success": 0.0,
492
+ "hidden_chart_oracle_utility": 0.2857772707939148,
493
+ "outcome_ptr": 0.0,
494
+ "predicted_margin": 0.3024309090323956,
495
+ "proposal_oracle_success": 0.0,
496
+ "proposal_oracle_utility": 0.007009605411440134,
497
+ "selected_candidate_index": 7.0,
498
+ "selected_success": 0.0,
499
+ "selected_success_gain_over_base": 0.0,
500
+ "selected_utility": 0.0069834389723837376,
501
+ "selected_utility_gain_over_base": -0.09585080808028579,
502
+ "selector_regret": 2.6166439056396484e-05,
503
+ "success_selector_gap": 0.0,
504
+ "success_support_gap": 0.0,
505
+ "support_gap": 0.27876766538247466,
506
+ "tau": 0.288910690801741
507
+ },
508
+ "141984615/train1": {
509
+ "base_success": 0.0,
510
+ "base_utility": 0.10283424705266953,
511
+ "coverage": 1.0,
512
+ "execute_generated": 1.0,
513
+ "fallback_rate": 0.0,
514
+ "hidden_chart_oracle_success": 0.0,
515
+ "hidden_chart_oracle_utility": 0.2857772707939148,
516
+ "outcome_ptr": 0.0,
517
+ "predicted_margin": 0.3649340282417526,
518
+ "proposal_oracle_success": 0.0,
519
+ "proposal_oracle_utility": 0.03353220224380493,
520
+ "selected_candidate_index": 2.0,
521
+ "selected_success": 0.0,
522
+ "selected_success_gain_over_base": 0.0,
523
+ "selected_utility": 0.0267853531986475,
524
+ "selected_utility_gain_over_base": -0.07604889385402203,
525
+ "selector_regret": 0.0067468490451574326,
526
+ "success_selector_gap": 0.0,
527
+ "success_support_gap": 0.0,
528
+ "support_gap": 0.25224506855010986,
529
+ "tau": 0.288910690801741
530
+ },
531
+ "141984615/train2": {
532
+ "base_success": 0.0,
533
+ "base_utility": 0.10283424705266953,
534
+ "coverage": 1.0,
535
+ "execute_generated": 1.0,
536
+ "fallback_rate": 0.0,
537
+ "hidden_chart_oracle_success": 0.0,
538
+ "hidden_chart_oracle_utility": 0.2857772707939148,
539
+ "outcome_ptr": 0.0,
540
+ "predicted_margin": 0.31585285572767974,
541
+ "proposal_oracle_success": 0.0,
542
+ "proposal_oracle_utility": 0.029999038204550743,
543
+ "selected_candidate_index": 3.0,
544
+ "selected_success": 0.0,
545
+ "selected_success_gain_over_base": 0.0,
546
+ "selected_utility": 0.022743364796042442,
547
+ "selected_utility_gain_over_base": -0.08009088225662708,
548
+ "selector_regret": 0.007255673408508301,
549
+ "success_selector_gap": 0.0,
550
+ "success_support_gap": 0.0,
551
+ "support_gap": 0.25577823258936405,
552
+ "tau": 0.288910690801741
553
+ },
554
+ "1461069713/train0": {
555
+ "base_success": 0.0,
556
+ "base_utility": 0.0939030647277832,
557
+ "coverage": 1.0,
558
+ "execute_generated": 1.0,
559
+ "fallback_rate": 0.0,
560
+ "hidden_chart_oracle_success": 0.0,
561
+ "hidden_chart_oracle_utility": 0.1593424528837204,
562
+ "outcome_ptr": 0.0,
563
+ "predicted_margin": 0.37438606797162827,
564
+ "proposal_oracle_success": 0.0,
565
+ "proposal_oracle_utility": 0.061755817383527756,
566
+ "selected_candidate_index": 3.0,
567
+ "selected_success": 0.0,
568
+ "selected_success_gain_over_base": 0.0,
569
+ "selected_utility": 0.03513211011886597,
570
+ "selected_utility_gain_over_base": -0.058770954608917236,
571
+ "selector_regret": 0.02662370726466179,
572
+ "success_selector_gap": 0.0,
573
+ "success_support_gap": 0.0,
574
+ "support_gap": 0.09758663550019264,
575
+ "tau": 0.288910690801741
576
+ },
577
+ "1461069713/train1": {
578
+ "base_success": 0.0,
579
+ "base_utility": 0.0939030647277832,
580
+ "coverage": 1.0,
581
+ "execute_generated": 1.0,
582
+ "fallback_rate": 0.0,
583
+ "hidden_chart_oracle_success": 0.0,
584
+ "hidden_chart_oracle_utility": 0.1593424528837204,
585
+ "outcome_ptr": 0.0,
586
+ "predicted_margin": 0.32269982991329793,
587
+ "proposal_oracle_success": 0.0,
588
+ "proposal_oracle_utility": 0.018597206100821495,
589
+ "selected_candidate_index": 1.0,
590
+ "selected_success": 0.0,
591
+ "selected_success_gain_over_base": 0.0,
592
+ "selected_utility": 0.007630705833435059,
593
+ "selected_utility_gain_over_base": -0.08627235889434814,
594
+ "selector_regret": 0.010966500267386436,
595
+ "success_selector_gap": 0.0,
596
+ "success_support_gap": 0.0,
597
+ "support_gap": 0.1407452467828989,
598
+ "tau": 0.288910690801741
599
+ },
600
+ "1461069713/train2": {
601
+ "base_success": 0.0,
602
+ "base_utility": 0.0939030647277832,
603
+ "coverage": 1.0,
604
+ "execute_generated": 1.0,
605
+ "fallback_rate": 0.0,
606
+ "hidden_chart_oracle_success": 0.0,
607
+ "hidden_chart_oracle_utility": 0.1593424528837204,
608
+ "outcome_ptr": 1.0,
609
+ "predicted_margin": 0.4254327828144852,
610
+ "proposal_oracle_success": 0.0,
611
+ "proposal_oracle_utility": 0.09607130289077759,
612
+ "selected_candidate_index": 1.0,
613
+ "selected_success": 0.0,
614
+ "selected_success_gain_over_base": 0.0,
615
+ "selected_utility": 0.0894298329949379,
616
+ "selected_utility_gain_over_base": -0.004473231732845306,
617
+ "selector_regret": 0.006641469895839691,
618
+ "success_selector_gap": 0.0,
619
+ "success_support_gap": 0.0,
620
+ "support_gap": 0.06327114999294281,
621
+ "tau": 0.288910690801741
622
+ },
623
+ "1480856810/train0": {
624
+ "base_success": 0.0,
625
+ "base_utility": 0.08703407645225525,
626
+ "coverage": 1.0,
627
+ "execute_generated": 1.0,
628
+ "fallback_rate": 0.0,
629
+ "hidden_chart_oracle_success": 0.0,
630
+ "hidden_chart_oracle_utility": 0.20948036015033722,
631
+ "outcome_ptr": 1.0,
632
+ "predicted_margin": 0.35506897376848157,
633
+ "proposal_oracle_success": 1.0,
634
+ "proposal_oracle_utility": 2.0,
635
+ "selected_candidate_index": 3.0,
636
+ "selected_success": 0.0,
637
+ "selected_success_gain_over_base": 0.0,
638
+ "selected_utility": 0.07618571817874908,
639
+ "selected_utility_gain_over_base": -0.010848358273506165,
640
+ "selector_regret": 1.923814281821251,
641
+ "success_selector_gap": 1.0,
642
+ "success_support_gap": 0.0,
643
+ "support_gap": 0.0,
644
+ "tau": 0.288910690801741
645
+ },
646
+ "1480856810/train1": {
647
+ "base_success": 0.0,
648
+ "base_utility": 0.08703407645225525,
649
+ "coverage": 1.0,
650
+ "execute_generated": 1.0,
651
+ "fallback_rate": 0.0,
652
+ "hidden_chart_oracle_success": 0.0,
653
+ "hidden_chart_oracle_utility": 0.20948036015033722,
654
+ "outcome_ptr": 1.0,
655
+ "predicted_margin": 0.4001798503700866,
656
+ "proposal_oracle_success": 1.0,
657
+ "proposal_oracle_utility": 2.0,
658
+ "selected_candidate_index": 0.0,
659
+ "selected_success": 1.0,
660
+ "selected_success_gain_over_base": 1.0,
661
+ "selected_utility": 2.0,
662
+ "selected_utility_gain_over_base": 1.9129659235477448,
663
+ "selector_regret": 0.0,
664
+ "success_selector_gap": 0.0,
665
+ "success_support_gap": 0.0,
666
+ "support_gap": 0.0,
667
+ "tau": 0.288910690801741
668
+ },
669
+ "1480856810/train2": {
670
+ "base_success": 0.0,
671
+ "base_utility": 0.08703407645225525,
672
+ "coverage": 1.0,
673
+ "execute_generated": 1.0,
674
+ "fallback_rate": 0.0,
675
+ "hidden_chart_oracle_success": 0.0,
676
+ "hidden_chart_oracle_utility": 0.20948036015033722,
677
+ "outcome_ptr": 1.0,
678
+ "predicted_margin": 0.36139635858026325,
679
+ "proposal_oracle_success": 1.0,
680
+ "proposal_oracle_utility": 2.0,
681
+ "selected_candidate_index": 2.0,
682
+ "selected_success": 1.0,
683
+ "selected_success_gain_over_base": 1.0,
684
+ "selected_utility": 2.0,
685
+ "selected_utility_gain_over_base": 1.9129659235477448,
686
+ "selector_regret": 0.0,
687
+ "success_selector_gap": 0.0,
688
+ "success_support_gap": 0.0,
689
+ "support_gap": 0.0,
690
+ "tau": 0.288910690801741
691
+ },
692
+ "1509980664/train0": {
693
+ "base_success": 0.0,
694
+ "base_utility": 0.8334100246429443,
695
+ "coverage": 1.0,
696
+ "execute_generated": 1.0,
697
+ "fallback_rate": 0.0,
698
+ "hidden_chart_oracle_success": 1.0,
699
+ "hidden_chart_oracle_utility": 2.0,
700
+ "outcome_ptr": 1.0,
701
+ "predicted_margin": 0.35070404819759693,
702
+ "proposal_oracle_success": 1.0,
703
+ "proposal_oracle_utility": 2.0,
704
+ "selected_candidate_index": 1.0,
705
+ "selected_success": 0.0,
706
+ "selected_success_gain_over_base": 0.0,
707
+ "selected_utility": 0.18938380479812622,
708
+ "selected_utility_gain_over_base": -0.6440262198448181,
709
+ "selector_regret": 1.8106161952018738,
710
+ "success_selector_gap": 1.0,
711
+ "success_support_gap": 0.0,
712
+ "support_gap": 0.0,
713
+ "tau": 0.288910690801741
714
+ },
715
+ "1509980664/train1": {
716
+ "base_success": 0.0,
717
+ "base_utility": 0.8334100246429443,
718
+ "coverage": 1.0,
719
+ "execute_generated": 1.0,
720
+ "fallback_rate": 0.0,
721
+ "hidden_chart_oracle_success": 1.0,
722
+ "hidden_chart_oracle_utility": 2.0,
723
+ "outcome_ptr": 1.0,
724
+ "predicted_margin": 0.365553744501998,
725
+ "proposal_oracle_success": 1.0,
726
+ "proposal_oracle_utility": 2.0,
727
+ "selected_candidate_index": 5.0,
728
+ "selected_success": 0.0,
729
+ "selected_success_gain_over_base": 0.0,
730
+ "selected_utility": 0.830393373966217,
731
+ "selected_utility_gain_over_base": -0.003016650676727295,
732
+ "selector_regret": 1.169606626033783,
733
+ "success_selector_gap": 1.0,
734
+ "success_support_gap": 0.0,
735
+ "support_gap": 0.0,
736
+ "tau": 0.288910690801741
737
+ },
738
+ "1509980664/train2": {
739
+ "base_success": 0.0,
740
+ "base_utility": 0.8334100246429443,
741
+ "coverage": 1.0,
742
+ "execute_generated": 1.0,
743
+ "fallback_rate": 0.0,
744
+ "hidden_chart_oracle_success": 1.0,
745
+ "hidden_chart_oracle_utility": 2.0,
746
+ "outcome_ptr": 0.0,
747
+ "predicted_margin": 0.3628723168371443,
748
+ "proposal_oracle_success": 0.0,
749
+ "proposal_oracle_utility": 0.813949465751648,
750
+ "selected_candidate_index": 5.0,
751
+ "selected_success": 0.0,
752
+ "selected_success_gain_over_base": 0.0,
753
+ "selected_utility": 0.17479392886161804,
754
+ "selected_utility_gain_over_base": -0.6586160957813263,
755
+ "selector_regret": 0.6391555368900299,
756
+ "success_selector_gap": 0.0,
757
+ "success_support_gap": 1.0,
758
+ "support_gap": 1.186050534248352,
759
+ "tau": 0.288910690801741
760
+ },
761
+ "1522272686/train0": {
762
+ "base_success": 0.0,
763
+ "base_utility": 0.5839591026306152,
764
+ "coverage": 1.0,
765
+ "execute_generated": 1.0,
766
+ "fallback_rate": 0.0,
767
+ "hidden_chart_oracle_success": 1.0,
768
+ "hidden_chart_oracle_utility": 2.0,
769
+ "outcome_ptr": 0.0,
770
+ "predicted_margin": 0.3768153803261049,
771
+ "proposal_oracle_success": 0.0,
772
+ "proposal_oracle_utility": 0.18486915528774261,
773
+ "selected_candidate_index": 3.0,
774
+ "selected_success": 0.0,
775
+ "selected_success_gain_over_base": 0.0,
776
+ "selected_utility": 0.15725429356098175,
777
+ "selected_utility_gain_over_base": -0.4267048090696335,
778
+ "selector_regret": 0.027614861726760864,
779
+ "success_selector_gap": 0.0,
780
+ "success_support_gap": 1.0,
781
+ "support_gap": 1.8151308447122574,
782
+ "tau": 0.288910690801741
783
+ },
784
+ "1522272686/train1": {
785
+ "base_success": 0.0,
786
+ "base_utility": 0.5839591026306152,
787
+ "coverage": 1.0,
788
+ "execute_generated": 1.0,
789
+ "fallback_rate": 0.0,
790
+ "hidden_chart_oracle_success": 1.0,
791
+ "hidden_chart_oracle_utility": 2.0,
792
+ "outcome_ptr": 0.0,
793
+ "predicted_margin": 0.39312268047586346,
794
+ "proposal_oracle_success": 0.0,
795
+ "proposal_oracle_utility": 0.07581864297389984,
796
+ "selected_candidate_index": 5.0,
797
+ "selected_success": 0.0,
798
+ "selected_success_gain_over_base": 0.0,
799
+ "selected_utility": 0.057952869683504105,
800
+ "selected_utility_gain_over_base": -0.5260062329471111,
801
+ "selector_regret": 0.017865773290395737,
802
+ "success_selector_gap": 0.0,
803
+ "success_support_gap": 1.0,
804
+ "support_gap": 1.9241813570261002,
805
+ "tau": 0.288910690801741
806
+ },
807
+ "1522272686/train2": {
808
+ "base_success": 0.0,
809
+ "base_utility": 0.5839591026306152,
810
+ "coverage": 1.0,
811
+ "execute_generated": 1.0,
812
+ "fallback_rate": 0.0,
813
+ "hidden_chart_oracle_success": 1.0,
814
+ "hidden_chart_oracle_utility": 2.0,
815
+ "outcome_ptr": 0.0,
816
+ "predicted_margin": 0.4326397429355978,
817
+ "proposal_oracle_success": 0.0,
818
+ "proposal_oracle_utility": 0.1605987697839737,
819
+ "selected_candidate_index": 4.0,
820
+ "selected_success": 0.0,
821
+ "selected_success_gain_over_base": 0.0,
822
+ "selected_utility": 0.12158188968896866,
823
+ "selected_utility_gain_over_base": -0.4623772129416466,
824
+ "selector_regret": 0.039016880095005035,
825
+ "success_selector_gap": 0.0,
826
+ "success_support_gap": 1.0,
827
+ "support_gap": 1.8394012302160263,
828
+ "tau": 0.288910690801741
829
+ },
830
+ "1549324805/train0": {
831
+ "base_success": 0.0,
832
+ "base_utility": 0.10451483726501465,
833
+ "coverage": 1.0,
834
+ "execute_generated": 1.0,
835
+ "fallback_rate": 0.0,
836
+ "hidden_chart_oracle_success": 0.0,
837
+ "hidden_chart_oracle_utility": 0.2974560856819153,
838
+ "outcome_ptr": 1.0,
839
+ "predicted_margin": 0.45895607477623623,
840
+ "proposal_oracle_success": 0.0,
841
+ "proposal_oracle_utility": 0.12235603481531143,
842
+ "selected_candidate_index": 0.0,
843
+ "selected_success": 0.0,
844
+ "selected_success_gain_over_base": 0.0,
845
+ "selected_utility": 0.12235603481531143,
846
+ "selected_utility_gain_over_base": 0.017841197550296783,
847
+ "selector_regret": 0.0,
848
+ "success_selector_gap": 0.0,
849
+ "success_support_gap": 0.0,
850
+ "support_gap": 0.17510005086660385,
851
+ "tau": 0.288910690801741
852
+ },
853
+ "1549324805/train1": {
854
+ "base_success": 0.0,
855
+ "base_utility": 0.10451483726501465,
856
+ "coverage": 1.0,
857
+ "execute_generated": 1.0,
858
+ "fallback_rate": 0.0,
859
+ "hidden_chart_oracle_success": 0.0,
860
+ "hidden_chart_oracle_utility": 0.2974560856819153,
861
+ "outcome_ptr": 1.0,
862
+ "predicted_margin": 0.468444969407059,
863
+ "proposal_oracle_success": 0.0,
864
+ "proposal_oracle_utility": 0.19118134677410126,
865
+ "selected_candidate_index": 0.0,
866
+ "selected_success": 0.0,
867
+ "selected_success_gain_over_base": 0.0,
868
+ "selected_utility": 0.19118134677410126,
869
+ "selected_utility_gain_over_base": 0.08666650950908661,
870
+ "selector_regret": 0.0,
871
+ "success_selector_gap": 0.0,
872
+ "success_support_gap": 0.0,
873
+ "support_gap": 0.10627473890781403,
874
+ "tau": 0.288910690801741
875
+ },
876
+ "1549324805/train2": {
877
+ "base_success": 0.0,
878
+ "base_utility": 0.10451483726501465,
879
+ "coverage": 1.0,
880
+ "execute_generated": 1.0,
881
+ "fallback_rate": 0.0,
882
+ "hidden_chart_oracle_success": 0.0,
883
+ "hidden_chart_oracle_utility": 0.2974560856819153,
884
+ "outcome_ptr": 0.0,
885
+ "predicted_margin": 0.45828710420858776,
886
+ "proposal_oracle_success": 0.0,
887
+ "proposal_oracle_utility": 0.10127037763595581,
888
+ "selected_candidate_index": 1.0,
889
+ "selected_success": 0.0,
890
+ "selected_success_gain_over_base": 0.0,
891
+ "selected_utility": 0.049150723963975906,
892
+ "selected_utility_gain_over_base": -0.05536411330103874,
893
+ "selector_regret": 0.052119653671979904,
894
+ "success_selector_gap": 0.0,
895
+ "success_support_gap": 0.0,
896
+ "support_gap": 0.19618570804595947,
897
+ "tau": 0.288910690801741
898
+ },
899
+ "1578990937/train0": {
900
+ "base_success": 0.0,
901
+ "base_utility": 0.82413649559021,
902
+ "coverage": 1.0,
903
+ "execute_generated": 1.0,
904
+ "fallback_rate": 0.0,
905
+ "hidden_chart_oracle_success": 1.0,
906
+ "hidden_chart_oracle_utility": 2.0,
907
+ "outcome_ptr": 1.0,
908
+ "predicted_margin": 0.4047529121633037,
909
+ "proposal_oracle_success": 1.0,
910
+ "proposal_oracle_utility": 2.0,
911
+ "selected_candidate_index": 5.0,
912
+ "selected_success": 0.0,
913
+ "selected_success_gain_over_base": 0.0,
914
+ "selected_utility": 0.20158150792121887,
915
+ "selected_utility_gain_over_base": -0.6225549876689911,
916
+ "selector_regret": 1.7984184920787811,
917
+ "success_selector_gap": 1.0,
918
+ "success_support_gap": 0.0,
919
+ "support_gap": 0.0,
920
+ "tau": 0.288910690801741
921
+ },
922
+ "1578990937/train1": {
923
+ "base_success": 0.0,
924
+ "base_utility": 0.82413649559021,
925
+ "coverage": 1.0,
926
+ "execute_generated": 1.0,
927
+ "fallback_rate": 0.0,
928
+ "hidden_chart_oracle_success": 1.0,
929
+ "hidden_chart_oracle_utility": 2.0,
930
+ "outcome_ptr": 1.0,
931
+ "predicted_margin": 0.35890637572307993,
932
+ "proposal_oracle_success": 1.0,
933
+ "proposal_oracle_utility": 2.0,
934
+ "selected_candidate_index": 1.0,
935
+ "selected_success": 1.0,
936
+ "selected_success_gain_over_base": 1.0,
937
+ "selected_utility": 2.0,
938
+ "selected_utility_gain_over_base": 1.17586350440979,
939
+ "selector_regret": 0.0,
940
+ "success_selector_gap": 0.0,
941
+ "success_support_gap": 0.0,
942
+ "support_gap": 0.0,
943
+ "tau": 0.288910690801741
944
+ },
945
+ "1578990937/train2": {
946
+ "base_success": 0.0,
947
+ "base_utility": 0.82413649559021,
948
+ "coverage": 1.0,
949
+ "execute_generated": 1.0,
950
+ "fallback_rate": 0.0,
951
+ "hidden_chart_oracle_success": 1.0,
952
+ "hidden_chart_oracle_utility": 2.0,
953
+ "outcome_ptr": 1.0,
954
+ "predicted_margin": 0.37963097790283945,
955
+ "proposal_oracle_success": 0.0,
956
+ "proposal_oracle_utility": 0.8332412242889404,
957
+ "selected_candidate_index": 2.0,
958
+ "selected_success": 0.0,
959
+ "selected_success_gain_over_base": 0.0,
960
+ "selected_utility": 0.8332412242889404,
961
+ "selected_utility_gain_over_base": 0.009104728698730469,
962
+ "selector_regret": 0.0,
963
+ "success_selector_gap": 0.0,
964
+ "success_support_gap": 1.0,
965
+ "support_gap": 1.1667587757110596,
966
+ "tau": 0.288910690801741
967
+ },
968
+ "1747178030/train0": {
969
+ "base_success": 1.0,
970
+ "base_utility": 2.0,
971
+ "coverage": 1.0,
972
+ "execute_generated": 1.0,
973
+ "fallback_rate": 0.0,
974
+ "hidden_chart_oracle_success": 1.0,
975
+ "hidden_chart_oracle_utility": 2.0,
976
+ "outcome_ptr": 0.0,
977
+ "predicted_margin": 0.38977296214707813,
978
+ "proposal_oracle_success": 1.0,
979
+ "proposal_oracle_utility": 2.0,
980
+ "selected_candidate_index": 1.0,
981
+ "selected_success": 1.0,
982
+ "selected_success_gain_over_base": 0.0,
983
+ "selected_utility": 2.0,
984
+ "selected_utility_gain_over_base": 0.0,
985
+ "selector_regret": 0.0,
986
+ "success_selector_gap": 0.0,
987
+ "success_support_gap": 0.0,
988
+ "support_gap": 0.0,
989
+ "tau": 0.288910690801741
990
+ },
991
+ "1747178030/train1": {
992
+ "base_success": 1.0,
993
+ "base_utility": 2.0,
994
+ "coverage": 1.0,
995
+ "execute_generated": 1.0,
996
+ "fallback_rate": 0.0,
997
+ "hidden_chart_oracle_success": 1.0,
998
+ "hidden_chart_oracle_utility": 2.0,
999
+ "outcome_ptr": 0.0,
1000
+ "predicted_margin": 0.347680836223678,
1001
+ "proposal_oracle_success": 1.0,
1002
+ "proposal_oracle_utility": 2.0,
1003
+ "selected_candidate_index": 1.0,
1004
+ "selected_success": 1.0,
1005
+ "selected_success_gain_over_base": 0.0,
1006
+ "selected_utility": 2.0,
1007
+ "selected_utility_gain_over_base": 0.0,
1008
+ "selector_regret": 0.0,
1009
+ "success_selector_gap": 0.0,
1010
+ "success_support_gap": 0.0,
1011
+ "support_gap": 0.0,
1012
+ "tau": 0.288910690801741
1013
+ },
1014
+ "1747178030/train2": {
1015
+ "base_success": 1.0,
1016
+ "base_utility": 2.0,
1017
+ "coverage": 1.0,
1018
+ "execute_generated": 1.0,
1019
+ "fallback_rate": 0.0,
1020
+ "hidden_chart_oracle_success": 1.0,
1021
+ "hidden_chart_oracle_utility": 2.0,
1022
+ "outcome_ptr": 0.0,
1023
+ "predicted_margin": 0.36537575470050154,
1024
+ "proposal_oracle_success": 1.0,
1025
+ "proposal_oracle_utility": 2.0,
1026
+ "selected_candidate_index": 5.0,
1027
+ "selected_success": 1.0,
1028
+ "selected_success_gain_over_base": 0.0,
1029
+ "selected_utility": 2.0,
1030
+ "selected_utility_gain_over_base": 0.0,
1031
+ "selector_regret": 0.0,
1032
+ "success_selector_gap": 0.0,
1033
+ "success_support_gap": 0.0,
1034
+ "support_gap": 0.0,
1035
+ "tau": 0.288910690801741
1036
+ },
1037
+ "1790188787/train0": {
1038
+ "base_success": 0.0,
1039
+ "base_utility": 0.12476746737957001,
1040
+ "coverage": 1.0,
1041
+ "execute_generated": 1.0,
1042
+ "fallback_rate": 0.0,
1043
+ "hidden_chart_oracle_success": 1.0,
1044
+ "hidden_chart_oracle_utility": 2.0,
1045
+ "outcome_ptr": 1.0,
1046
+ "predicted_margin": 0.3153898365232336,
1047
+ "proposal_oracle_success": 1.0,
1048
+ "proposal_oracle_utility": 2.0,
1049
+ "selected_candidate_index": 4.0,
1050
+ "selected_success": 0.0,
1051
+ "selected_success_gain_over_base": 0.0,
1052
+ "selected_utility": 0.07329326868057251,
1053
+ "selected_utility_gain_over_base": -0.0514741986989975,
1054
+ "selector_regret": 1.9267067313194275,
1055
+ "success_selector_gap": 1.0,
1056
+ "success_support_gap": 0.0,
1057
+ "support_gap": 0.0,
1058
+ "tau": 0.288910690801741
1059
+ },
1060
+ "1790188787/train1": {
1061
+ "base_success": 0.0,
1062
+ "base_utility": 0.12476746737957001,
1063
+ "coverage": 0.0,
1064
+ "execute_generated": 0.0,
1065
+ "fallback_rate": 1.0,
1066
+ "hidden_chart_oracle_success": 1.0,
1067
+ "hidden_chart_oracle_utility": 2.0,
1068
+ "outcome_ptr": 0.0,
1069
+ "predicted_margin": 0.26995857568215537,
1070
+ "proposal_oracle_success": 0.0,
1071
+ "proposal_oracle_utility": 0.09691467881202698,
1072
+ "selected_candidate_index": 1.0,
1073
+ "selected_success": 0.0,
1074
+ "selected_success_gain_over_base": 0.0,
1075
+ "selected_utility": 0.12476746737957001,
1076
+ "selected_utility_gain_over_base": 0.0,
1077
+ "selector_regret": 0.0,
1078
+ "success_selector_gap": 0.0,
1079
+ "success_support_gap": 1.0,
1080
+ "support_gap": 1.903085321187973,
1081
+ "tau": 0.288910690801741
1082
+ },
1083
+ "1790188787/train2": {
1084
+ "base_success": 0.0,
1085
+ "base_utility": 0.12476746737957001,
1086
+ "coverage": 1.0,
1087
+ "execute_generated": 1.0,
1088
+ "fallback_rate": 0.0,
1089
+ "hidden_chart_oracle_success": 1.0,
1090
+ "hidden_chart_oracle_utility": 2.0,
1091
+ "outcome_ptr": 1.0,
1092
+ "predicted_margin": 0.33260725289666326,
1093
+ "proposal_oracle_success": 0.0,
1094
+ "proposal_oracle_utility": 0.13471829891204834,
1095
+ "selected_candidate_index": 2.0,
1096
+ "selected_success": 0.0,
1097
+ "selected_success_gain_over_base": 0.0,
1098
+ "selected_utility": 0.09578777849674225,
1099
+ "selected_utility_gain_over_base": -0.02897968888282776,
1100
+ "selector_regret": 0.03893052041530609,
1101
+ "success_selector_gap": 0.0,
1102
+ "success_support_gap": 1.0,
1103
+ "support_gap": 1.8652817010879517,
1104
+ "tau": 0.288910690801741
1105
+ },
1106
+ "1822210471/train0": {
1107
+ "base_success": 0.0,
1108
+ "base_utility": 0.07191530615091324,
1109
+ "coverage": 0.0,
1110
+ "execute_generated": 0.0,
1111
+ "fallback_rate": 1.0,
1112
+ "hidden_chart_oracle_success": 1.0,
1113
+ "hidden_chart_oracle_utility": 2.0,
1114
+ "outcome_ptr": 1.0,
1115
+ "predicted_margin": 0.24984280166583867,
1116
+ "proposal_oracle_success": 1.0,
1117
+ "proposal_oracle_utility": 2.0,
1118
+ "selected_candidate_index": 4.0,
1119
+ "selected_success": 0.0,
1120
+ "selected_success_gain_over_base": 0.0,
1121
+ "selected_utility": 0.07191530615091324,
1122
+ "selected_utility_gain_over_base": 0.0,
1123
+ "selector_regret": 1.9280846938490868,
1124
+ "success_selector_gap": 1.0,
1125
+ "success_support_gap": 0.0,
1126
+ "support_gap": 0.0,
1127
+ "tau": 0.288910690801741
1128
+ },
1129
+ "1822210471/train1": {
1130
+ "base_success": 0.0,
1131
+ "base_utility": 0.07191530615091324,
1132
+ "coverage": 0.0,
1133
+ "execute_generated": 0.0,
1134
+ "fallback_rate": 1.0,
1135
+ "hidden_chart_oracle_success": 1.0,
1136
+ "hidden_chart_oracle_utility": 2.0,
1137
+ "outcome_ptr": 1.0,
1138
+ "predicted_margin": 0.2582953317465545,
1139
+ "proposal_oracle_success": 1.0,
1140
+ "proposal_oracle_utility": 2.0,
1141
+ "selected_candidate_index": 3.0,
1142
+ "selected_success": 0.0,
1143
+ "selected_success_gain_over_base": 0.0,
1144
+ "selected_utility": 0.07191530615091324,
1145
+ "selected_utility_gain_over_base": 0.0,
1146
+ "selector_regret": 1.9280846938490868,
1147
+ "success_selector_gap": 1.0,
1148
+ "success_support_gap": 0.0,
1149
+ "support_gap": 0.0,
1150
+ "tau": 0.288910690801741
1151
+ },
1152
+ "1822210471/train2": {
1153
+ "base_success": 0.0,
1154
+ "base_utility": 0.07191530615091324,
1155
+ "coverage": 0.0,
1156
+ "execute_generated": 0.0,
1157
+ "fallback_rate": 1.0,
1158
+ "hidden_chart_oracle_success": 1.0,
1159
+ "hidden_chart_oracle_utility": 2.0,
1160
+ "outcome_ptr": 1.0,
1161
+ "predicted_margin": 0.2796914476829909,
1162
+ "proposal_oracle_success": 1.0,
1163
+ "proposal_oracle_utility": 2.0,
1164
+ "selected_candidate_index": 5.0,
1165
+ "selected_success": 0.0,
1166
+ "selected_success_gain_over_base": 0.0,
1167
+ "selected_utility": 0.07191530615091324,
1168
+ "selected_utility_gain_over_base": 0.0,
1169
+ "selector_regret": 1.9280846938490868,
1170
+ "success_selector_gap": 1.0,
1171
+ "success_support_gap": 0.0,
1172
+ "support_gap": 0.0,
1173
+ "tau": 0.288910690801741
1174
+ },
1175
+ "1845489013/train0": {
1176
+ "base_success": 1.0,
1177
+ "base_utility": 2.0,
1178
+ "coverage": 1.0,
1179
+ "execute_generated": 1.0,
1180
+ "fallback_rate": 0.0,
1181
+ "hidden_chart_oracle_success": 1.0,
1182
+ "hidden_chart_oracle_utility": 2.0,
1183
+ "outcome_ptr": 0.0,
1184
+ "predicted_margin": 0.3447311069211071,
1185
+ "proposal_oracle_success": 1.0,
1186
+ "proposal_oracle_utility": 2.0,
1187
+ "selected_candidate_index": 3.0,
1188
+ "selected_success": 1.0,
1189
+ "selected_success_gain_over_base": 0.0,
1190
+ "selected_utility": 2.0,
1191
+ "selected_utility_gain_over_base": 0.0,
1192
+ "selector_regret": 0.0,
1193
+ "success_selector_gap": 0.0,
1194
+ "success_support_gap": 0.0,
1195
+ "support_gap": 0.0,
1196
+ "tau": 0.288910690801741
1197
+ },
1198
+ "1845489013/train1": {
1199
+ "base_success": 1.0,
1200
+ "base_utility": 2.0,
1201
+ "coverage": 1.0,
1202
+ "execute_generated": 1.0,
1203
+ "fallback_rate": 0.0,
1204
+ "hidden_chart_oracle_success": 1.0,
1205
+ "hidden_chart_oracle_utility": 2.0,
1206
+ "outcome_ptr": 0.0,
1207
+ "predicted_margin": 0.34570350245731873,
1208
+ "proposal_oracle_success": 1.0,
1209
+ "proposal_oracle_utility": 2.0,
1210
+ "selected_candidate_index": 1.0,
1211
+ "selected_success": 1.0,
1212
+ "selected_success_gain_over_base": 0.0,
1213
+ "selected_utility": 2.0,
1214
+ "selected_utility_gain_over_base": 0.0,
1215
+ "selector_regret": 0.0,
1216
+ "success_selector_gap": 0.0,
1217
+ "success_support_gap": 0.0,
1218
+ "support_gap": 0.0,
1219
+ "tau": 0.288910690801741
1220
+ },
1221
+ "1845489013/train2": {
1222
+ "base_success": 1.0,
1223
+ "base_utility": 2.0,
1224
+ "coverage": 1.0,
1225
+ "execute_generated": 1.0,
1226
+ "fallback_rate": 0.0,
1227
+ "hidden_chart_oracle_success": 1.0,
1228
+ "hidden_chart_oracle_utility": 2.0,
1229
+ "outcome_ptr": 0.0,
1230
+ "predicted_margin": 0.33555214554269086,
1231
+ "proposal_oracle_success": 1.0,
1232
+ "proposal_oracle_utility": 2.0,
1233
+ "selected_candidate_index": 1.0,
1234
+ "selected_success": 1.0,
1235
+ "selected_success_gain_over_base": 0.0,
1236
+ "selected_utility": 2.0,
1237
+ "selected_utility_gain_over_base": 0.0,
1238
+ "selector_regret": 0.0,
1239
+ "success_selector_gap": 0.0,
1240
+ "success_support_gap": 0.0,
1241
+ "support_gap": 0.0,
1242
+ "tau": 0.288910690801741
1243
+ },
1244
+ "1999022574/train0": {
1245
+ "base_success": 1.0,
1246
+ "base_utility": 2.0,
1247
+ "coverage": 1.0,
1248
+ "execute_generated": 1.0,
1249
+ "fallback_rate": 0.0,
1250
+ "hidden_chart_oracle_success": 1.0,
1251
+ "hidden_chart_oracle_utility": 2.0,
1252
+ "outcome_ptr": 0.0,
1253
+ "predicted_margin": 0.3211068056671615,
1254
+ "proposal_oracle_success": 1.0,
1255
+ "proposal_oracle_utility": 2.0,
1256
+ "selected_candidate_index": 4.0,
1257
+ "selected_success": 0.0,
1258
+ "selected_success_gain_over_base": -1.0,
1259
+ "selected_utility": 0.1689002811908722,
1260
+ "selected_utility_gain_over_base": -1.8310997188091278,
1261
+ "selector_regret": 1.8310997188091278,
1262
+ "success_selector_gap": 1.0,
1263
+ "success_support_gap": 0.0,
1264
+ "support_gap": 0.0,
1265
+ "tau": 0.288910690801741
1266
+ },
1267
+ "1999022574/train1": {
1268
+ "base_success": 1.0,
1269
+ "base_utility": 2.0,
1270
+ "coverage": 1.0,
1271
+ "execute_generated": 1.0,
1272
+ "fallback_rate": 0.0,
1273
+ "hidden_chart_oracle_success": 1.0,
1274
+ "hidden_chart_oracle_utility": 2.0,
1275
+ "outcome_ptr": 0.0,
1276
+ "predicted_margin": 0.3156294299353899,
1277
+ "proposal_oracle_success": 1.0,
1278
+ "proposal_oracle_utility": 2.0,
1279
+ "selected_candidate_index": 4.0,
1280
+ "selected_success": 1.0,
1281
+ "selected_success_gain_over_base": 0.0,
1282
+ "selected_utility": 2.0,
1283
+ "selected_utility_gain_over_base": 0.0,
1284
+ "selector_regret": 0.0,
1285
+ "success_selector_gap": 0.0,
1286
+ "success_support_gap": 0.0,
1287
+ "support_gap": 0.0,
1288
+ "tau": 0.288910690801741
1289
+ },
1290
+ "1999022574/train2": {
1291
+ "base_success": 1.0,
1292
+ "base_utility": 2.0,
1293
+ "coverage": 1.0,
1294
+ "execute_generated": 1.0,
1295
+ "fallback_rate": 0.0,
1296
+ "hidden_chart_oracle_success": 1.0,
1297
+ "hidden_chart_oracle_utility": 2.0,
1298
+ "outcome_ptr": 0.0,
1299
+ "predicted_margin": 0.32337789122159327,
1300
+ "proposal_oracle_success": 0.0,
1301
+ "proposal_oracle_utility": 0.8125013113021851,
1302
+ "selected_candidate_index": 4.0,
1303
+ "selected_success": 0.0,
1304
+ "selected_success_gain_over_base": -1.0,
1305
+ "selected_utility": 0.16747623682022095,
1306
+ "selected_utility_gain_over_base": -1.832523763179779,
1307
+ "selector_regret": 0.6450250744819641,
1308
+ "success_selector_gap": 0.0,
1309
+ "success_support_gap": 1.0,
1310
+ "support_gap": 1.187498688697815,
1311
+ "tau": 0.288910690801741
1312
+ },
1313
+ "2012211287/train0": {
1314
+ "base_success": 0.0,
1315
+ "base_utility": 0.7236053347587585,
1316
+ "coverage": 1.0,
1317
+ "execute_generated": 1.0,
1318
+ "fallback_rate": 0.0,
1319
+ "hidden_chart_oracle_success": 1.0,
1320
+ "hidden_chart_oracle_utility": 2.0,
1321
+ "outcome_ptr": 1.0,
1322
+ "predicted_margin": 0.31307179175386435,
1323
+ "proposal_oracle_success": 1.0,
1324
+ "proposal_oracle_utility": 2.0,
1325
+ "selected_candidate_index": 5.0,
1326
+ "selected_success": 0.0,
1327
+ "selected_success_gain_over_base": 0.0,
1328
+ "selected_utility": 0.6998441815376282,
1329
+ "selected_utility_gain_over_base": -0.02376115322113037,
1330
+ "selector_regret": 1.3001558184623718,
1331
+ "success_selector_gap": 1.0,
1332
+ "success_support_gap": 0.0,
1333
+ "support_gap": 0.0,
1334
+ "tau": 0.288910690801741
1335
+ },
1336
+ "2012211287/train1": {
1337
+ "base_success": 0.0,
1338
+ "base_utility": 0.7236053347587585,
1339
+ "coverage": 0.0,
1340
+ "execute_generated": 0.0,
1341
+ "fallback_rate": 1.0,
1342
+ "hidden_chart_oracle_success": 1.0,
1343
+ "hidden_chart_oracle_utility": 2.0,
1344
+ "outcome_ptr": 1.0,
1345
+ "predicted_margin": 0.2790826163422787,
1346
+ "proposal_oracle_success": 1.0,
1347
+ "proposal_oracle_utility": 2.0,
1348
+ "selected_candidate_index": 1.0,
1349
+ "selected_success": 0.0,
1350
+ "selected_success_gain_over_base": 0.0,
1351
+ "selected_utility": 0.7236053347587585,
1352
+ "selected_utility_gain_over_base": 0.0,
1353
+ "selector_regret": 1.2763946652412415,
1354
+ "success_selector_gap": 1.0,
1355
+ "success_support_gap": 0.0,
1356
+ "support_gap": 0.0,
1357
+ "tau": 0.288910690801741
1358
+ },
1359
+ "2012211287/train2": {
1360
+ "base_success": 0.0,
1361
+ "base_utility": 0.7236053347587585,
1362
+ "coverage": 1.0,
1363
+ "execute_generated": 1.0,
1364
+ "fallback_rate": 0.0,
1365
+ "hidden_chart_oracle_success": 1.0,
1366
+ "hidden_chart_oracle_utility": 2.0,
1367
+ "outcome_ptr": 1.0,
1368
+ "predicted_margin": 0.3316486280164904,
1369
+ "proposal_oracle_success": 1.0,
1370
+ "proposal_oracle_utility": 2.0,
1371
+ "selected_candidate_index": 4.0,
1372
+ "selected_success": 1.0,
1373
+ "selected_success_gain_over_base": 1.0,
1374
+ "selected_utility": 2.0,
1375
+ "selected_utility_gain_over_base": 1.2763946652412415,
1376
+ "selector_regret": 0.0,
1377
+ "success_selector_gap": 0.0,
1378
+ "success_support_gap": 0.0,
1379
+ "support_gap": 0.0,
1380
+ "tau": 0.288910690801741
1381
+ },
1382
+ "2038594191/train0": {
1383
+ "base_success": 1.0,
1384
+ "base_utility": 2.0,
1385
+ "coverage": 1.0,
1386
+ "execute_generated": 1.0,
1387
+ "fallback_rate": 0.0,
1388
+ "hidden_chart_oracle_success": 1.0,
1389
+ "hidden_chart_oracle_utility": 2.0,
1390
+ "outcome_ptr": 0.0,
1391
+ "predicted_margin": 0.3096675011178494,
1392
+ "proposal_oracle_success": 1.0,
1393
+ "proposal_oracle_utility": 2.0,
1394
+ "selected_candidate_index": 3.0,
1395
+ "selected_success": 1.0,
1396
+ "selected_success_gain_over_base": 0.0,
1397
+ "selected_utility": 2.0,
1398
+ "selected_utility_gain_over_base": 0.0,
1399
+ "selector_regret": 0.0,
1400
+ "success_selector_gap": 0.0,
1401
+ "success_support_gap": 0.0,
1402
+ "support_gap": 0.0,
1403
+ "tau": 0.288910690801741
1404
+ },
1405
+ "2038594191/train1": {
1406
+ "base_success": 1.0,
1407
+ "base_utility": 2.0,
1408
+ "coverage": 1.0,
1409
+ "execute_generated": 1.0,
1410
+ "fallback_rate": 0.0,
1411
+ "hidden_chart_oracle_success": 1.0,
1412
+ "hidden_chart_oracle_utility": 2.0,
1413
+ "outcome_ptr": 0.0,
1414
+ "predicted_margin": 0.2970325154837386,
1415
+ "proposal_oracle_success": 1.0,
1416
+ "proposal_oracle_utility": 2.0,
1417
+ "selected_candidate_index": 0.0,
1418
+ "selected_success": 0.0,
1419
+ "selected_success_gain_over_base": -1.0,
1420
+ "selected_utility": 0.037331562489271164,
1421
+ "selected_utility_gain_over_base": -1.9626684375107288,
1422
+ "selector_regret": 1.9626684375107288,
1423
+ "success_selector_gap": 1.0,
1424
+ "success_support_gap": 0.0,
1425
+ "support_gap": 0.0,
1426
+ "tau": 0.288910690801741
1427
+ },
1428
+ "2038594191/train2": {
1429
+ "base_success": 1.0,
1430
+ "base_utility": 2.0,
1431
+ "coverage": 1.0,
1432
+ "execute_generated": 1.0,
1433
+ "fallback_rate": 0.0,
1434
+ "hidden_chart_oracle_success": 1.0,
1435
+ "hidden_chart_oracle_utility": 2.0,
1436
+ "outcome_ptr": 0.0,
1437
+ "predicted_margin": 0.3443522704819967,
1438
+ "proposal_oracle_success": 1.0,
1439
+ "proposal_oracle_utility": 2.0,
1440
+ "selected_candidate_index": 2.0,
1441
+ "selected_success": 1.0,
1442
+ "selected_success_gain_over_base": 0.0,
1443
+ "selected_utility": 2.0,
1444
+ "selected_utility_gain_over_base": 0.0,
1445
+ "selector_regret": 0.0,
1446
+ "success_selector_gap": 0.0,
1447
+ "success_support_gap": 0.0,
1448
+ "support_gap": 0.0,
1449
+ "tau": 0.288910690801741
1450
+ },
1451
+ "2054899126/train0": {
1452
+ "base_success": 1.0,
1453
+ "base_utility": 2.0,
1454
+ "coverage": 0.5,
1455
+ "execute_generated": 0.5,
1456
+ "fallback_rate": 0.5,
1457
+ "hidden_chart_oracle_success": 1.0,
1458
+ "hidden_chart_oracle_utility": 2.0,
1459
+ "outcome_ptr": 0.0,
1460
+ "predicted_margin": 0.34626491665062,
1461
+ "proposal_oracle_success": 0.5,
1462
+ "proposal_oracle_utility": 1.0882058069109917,
1463
+ "selected_candidate_index": 4.5,
1464
+ "selected_success": 1.0,
1465
+ "selected_success_gain_over_base": 0.0,
1466
+ "selected_utility": 2.0,
1467
+ "selected_utility_gain_over_base": 0.0,
1468
+ "selector_regret": 0.0,
1469
+ "success_selector_gap": 0.0,
1470
+ "success_support_gap": 0.5,
1471
+ "support_gap": 0.9117941930890083,
1472
+ "tau": 0.288910690801741
1473
+ },
1474
+ "2054899126/train1": {
1475
+ "base_success": 1.0,
1476
+ "base_utility": 2.0,
1477
+ "coverage": 0.5,
1478
+ "execute_generated": 0.5,
1479
+ "fallback_rate": 0.5,
1480
+ "hidden_chart_oracle_success": 1.0,
1481
+ "hidden_chart_oracle_utility": 2.0,
1482
+ "outcome_ptr": 0.0,
1483
+ "predicted_margin": 0.3885488271124979,
1484
+ "proposal_oracle_success": 0.5,
1485
+ "proposal_oracle_utility": 1.0847108364105225,
1486
+ "selected_candidate_index": 3.0,
1487
+ "selected_success": 1.0,
1488
+ "selected_success_gain_over_base": 0.0,
1489
+ "selected_utility": 2.0,
1490
+ "selected_utility_gain_over_base": 0.0,
1491
+ "selector_regret": 0.0,
1492
+ "success_selector_gap": 0.0,
1493
+ "success_support_gap": 0.5,
1494
+ "support_gap": 0.9152891635894775,
1495
+ "tau": 0.288910690801741
1496
+ },
1497
+ "2054899126/train2": {
1498
+ "base_success": 1.0,
1499
+ "base_utility": 2.0,
1500
+ "coverage": 0.5,
1501
+ "execute_generated": 0.5,
1502
+ "fallback_rate": 0.5,
1503
+ "hidden_chart_oracle_success": 1.0,
1504
+ "hidden_chart_oracle_utility": 2.0,
1505
+ "outcome_ptr": 0.0,
1506
+ "predicted_margin": 0.3010287298448785,
1507
+ "proposal_oracle_success": 0.5,
1508
+ "proposal_oracle_utility": 1.047328144311905,
1509
+ "selected_candidate_index": 0.0,
1510
+ "selected_success": 1.0,
1511
+ "selected_success_gain_over_base": 0.0,
1512
+ "selected_utility": 2.0,
1513
+ "selected_utility_gain_over_base": 0.0,
1514
+ "selector_regret": 0.0,
1515
+ "success_selector_gap": 0.0,
1516
+ "success_support_gap": 0.5,
1517
+ "support_gap": 0.9526718556880951,
1518
+ "tau": 0.288910690801741
1519
+ },
1520
+ "2061635888/train0": {
1521
+ "base_success": 0.0,
1522
+ "base_utility": 0.05225050449371338,
1523
+ "coverage": 1.0,
1524
+ "execute_generated": 1.0,
1525
+ "fallback_rate": 0.0,
1526
+ "hidden_chart_oracle_success": 1.0,
1527
+ "hidden_chart_oracle_utility": 2.0,
1528
+ "outcome_ptr": 1.0,
1529
+ "predicted_margin": 0.30736542116235854,
1530
+ "proposal_oracle_success": 0.0,
1531
+ "proposal_oracle_utility": 0.0980667769908905,
1532
+ "selected_candidate_index": 3.0,
1533
+ "selected_success": 0.0,
1534
+ "selected_success_gain_over_base": 0.0,
1535
+ "selected_utility": 0.09766748547554016,
1536
+ "selected_utility_gain_over_base": 0.04541698098182678,
1537
+ "selector_regret": 0.0003992915153503418,
1538
+ "success_selector_gap": 0.0,
1539
+ "success_support_gap": 1.0,
1540
+ "support_gap": 1.9019332230091095,
1541
+ "tau": 0.288910690801741
1542
+ },
1543
+ "2061635888/train1": {
1544
+ "base_success": 0.0,
1545
+ "base_utility": 0.05225050449371338,
1546
+ "coverage": 1.0,
1547
+ "execute_generated": 1.0,
1548
+ "fallback_rate": 0.0,
1549
+ "hidden_chart_oracle_success": 1.0,
1550
+ "hidden_chart_oracle_utility": 2.0,
1551
+ "outcome_ptr": 1.0,
1552
+ "predicted_margin": 0.3243745362125768,
1553
+ "proposal_oracle_success": 0.0,
1554
+ "proposal_oracle_utility": 0.10504114627838135,
1555
+ "selected_candidate_index": 2.0,
1556
+ "selected_success": 0.0,
1557
+ "selected_success_gain_over_base": 0.0,
1558
+ "selected_utility": 0.06736154854297638,
1559
+ "selected_utility_gain_over_base": 0.015111044049263,
1560
+ "selector_regret": 0.03767959773540497,
1561
+ "success_selector_gap": 0.0,
1562
+ "success_support_gap": 1.0,
1563
+ "support_gap": 1.8949588537216187,
1564
+ "tau": 0.288910690801741
1565
+ },
1566
+ "2061635888/train2": {
1567
+ "base_success": 0.0,
1568
+ "base_utility": 0.05225050449371338,
1569
+ "coverage": 1.0,
1570
+ "execute_generated": 1.0,
1571
+ "fallback_rate": 0.0,
1572
+ "hidden_chart_oracle_success": 1.0,
1573
+ "hidden_chart_oracle_utility": 2.0,
1574
+ "outcome_ptr": 1.0,
1575
+ "predicted_margin": 0.32257244395689183,
1576
+ "proposal_oracle_success": 0.0,
1577
+ "proposal_oracle_utility": 0.1095939576625824,
1578
+ "selected_candidate_index": 4.0,
1579
+ "selected_success": 0.0,
1580
+ "selected_success_gain_over_base": 0.0,
1581
+ "selected_utility": 0.08809271454811096,
1582
+ "selected_utility_gain_over_base": 0.03584221005439758,
1583
+ "selector_regret": 0.021501243114471436,
1584
+ "success_selector_gap": 0.0,
1585
+ "success_support_gap": 1.0,
1586
+ "support_gap": 1.8904060423374176,
1587
+ "tau": 0.288910690801741
1588
+ },
1589
+ "211245842/train0": {
1590
+ "base_success": 0.0,
1591
+ "base_utility": 0.19090841710567474,
1592
+ "coverage": 1.0,
1593
+ "execute_generated": 1.0,
1594
+ "fallback_rate": 0.0,
1595
+ "hidden_chart_oracle_success": 1.0,
1596
+ "hidden_chart_oracle_utility": 2.0,
1597
+ "outcome_ptr": 1.0,
1598
+ "predicted_margin": 0.3668623343405369,
1599
+ "proposal_oracle_success": 1.0,
1600
+ "proposal_oracle_utility": 2.0,
1601
+ "selected_candidate_index": 1.0,
1602
+ "selected_success": 0.0,
1603
+ "selected_success_gain_over_base": 0.0,
1604
+ "selected_utility": 0.12450027465820312,
1605
+ "selected_utility_gain_over_base": -0.06640814244747162,
1606
+ "selector_regret": 1.8754997253417969,
1607
+ "success_selector_gap": 1.0,
1608
+ "success_support_gap": 0.0,
1609
+ "support_gap": 0.0,
1610
+ "tau": 0.288910690801741
1611
+ },
1612
+ "211245842/train1": {
1613
+ "base_success": 0.0,
1614
+ "base_utility": 0.19090841710567474,
1615
+ "coverage": 1.0,
1616
+ "execute_generated": 1.0,
1617
+ "fallback_rate": 0.0,
1618
+ "hidden_chart_oracle_success": 1.0,
1619
+ "hidden_chart_oracle_utility": 2.0,
1620
+ "outcome_ptr": 1.0,
1621
+ "predicted_margin": 0.3191332662791156,
1622
+ "proposal_oracle_success": 0.0,
1623
+ "proposal_oracle_utility": 0.21834009885787964,
1624
+ "selected_candidate_index": 4.0,
1625
+ "selected_success": 0.0,
1626
+ "selected_success_gain_over_base": 0.0,
1627
+ "selected_utility": 0.17151470482349396,
1628
+ "selected_utility_gain_over_base": -0.019393712282180786,
1629
+ "selector_regret": 0.04682539403438568,
1630
+ "success_selector_gap": 0.0,
1631
+ "success_support_gap": 1.0,
1632
+ "support_gap": 1.7816599011421204,
1633
+ "tau": 0.288910690801741
1634
+ },
1635
+ "211245842/train2": {
1636
+ "base_success": 0.0,
1637
+ "base_utility": 0.19090841710567474,
1638
+ "coverage": 1.0,
1639
+ "execute_generated": 1.0,
1640
+ "fallback_rate": 0.0,
1641
+ "hidden_chart_oracle_success": 1.0,
1642
+ "hidden_chart_oracle_utility": 2.0,
1643
+ "outcome_ptr": 1.0,
1644
+ "predicted_margin": 0.37242923654463844,
1645
+ "proposal_oracle_success": 1.0,
1646
+ "proposal_oracle_utility": 2.0,
1647
+ "selected_candidate_index": 3.0,
1648
+ "selected_success": 0.0,
1649
+ "selected_success_gain_over_base": 0.0,
1650
+ "selected_utility": 0.2108742892742157,
1651
+ "selected_utility_gain_over_base": 0.019965872168540955,
1652
+ "selector_regret": 1.7891257107257843,
1653
+ "success_selector_gap": 1.0,
1654
+ "success_support_gap": 0.0,
1655
+ "support_gap": 0.0,
1656
+ "tau": 0.288910690801741
1657
+ },
1658
+ "278008524/train0": {
1659
+ "base_success": 0.0,
1660
+ "base_utility": 0.042734989896416664,
1661
+ "coverage": 0.5,
1662
+ "execute_generated": 0.5,
1663
+ "fallback_rate": 0.5,
1664
+ "hidden_chart_oracle_success": 0.5,
1665
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1666
+ "outcome_ptr": 1.0,
1667
+ "predicted_margin": 0.2793097080756189,
1668
+ "proposal_oracle_success": 0.5,
1669
+ "proposal_oracle_utility": 1.0255221128463745,
1670
+ "selected_candidate_index": 5.5,
1671
+ "selected_success": 0.0,
1672
+ "selected_success_gain_over_base": 0.0,
1673
+ "selected_utility": 0.0425200667232275,
1674
+ "selected_utility_gain_over_base": -0.0002149231731891632,
1675
+ "selector_regret": 0.983002046123147,
1676
+ "success_selector_gap": 0.5,
1677
+ "success_support_gap": 0.0,
1678
+ "support_gap": 0.08838652819395065,
1679
+ "tau": 0.288910690801741
1680
+ },
1681
+ "278008524/train1": {
1682
+ "base_success": 0.0,
1683
+ "base_utility": 0.042734989896416664,
1684
+ "coverage": 0.5,
1685
+ "execute_generated": 0.5,
1686
+ "fallback_rate": 0.5,
1687
+ "hidden_chart_oracle_success": 0.5,
1688
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1689
+ "outcome_ptr": 1.0,
1690
+ "predicted_margin": 0.31823423926837713,
1691
+ "proposal_oracle_success": 0.5,
1692
+ "proposal_oracle_utility": 1.0421672649681568,
1693
+ "selected_candidate_index": 3.5,
1694
+ "selected_success": 0.0,
1695
+ "selected_success_gain_over_base": 0.0,
1696
+ "selected_utility": 0.055249134078621864,
1697
+ "selected_utility_gain_over_base": 0.0125141441822052,
1698
+ "selector_regret": 0.986918130889535,
1699
+ "success_selector_gap": 0.5,
1700
+ "success_support_gap": 0.0,
1701
+ "support_gap": 0.07174137607216835,
1702
+ "tau": 0.288910690801741
1703
+ },
1704
+ "278008524/train2": {
1705
+ "base_success": 0.0,
1706
+ "base_utility": 0.042734989896416664,
1707
+ "coverage": 1.0,
1708
+ "execute_generated": 1.0,
1709
+ "fallback_rate": 0.0,
1710
+ "hidden_chart_oracle_success": 0.5,
1711
+ "hidden_chart_oracle_utility": 1.1139086410403252,
1712
+ "outcome_ptr": 1.0,
1713
+ "predicted_margin": 0.31496573660246774,
1714
+ "proposal_oracle_success": 0.5,
1715
+ "proposal_oracle_utility": 1.1361673325300217,
1716
+ "selected_candidate_index": 3.0,
1717
+ "selected_success": 0.0,
1718
+ "selected_success_gain_over_base": 0.0,
1719
+ "selected_utility": 0.1634208019822836,
1720
+ "selected_utility_gain_over_base": 0.12068581208586693,
1721
+ "selector_regret": 0.9727465305477381,
1722
+ "success_selector_gap": 0.5,
1723
+ "success_support_gap": 0.0,
1724
+ "support_gap": 0.0,
1725
+ "tau": 0.288910690801741
1726
+ },
1727
+ "307068879/train0": {
1728
+ "base_success": 0.0,
1729
+ "base_utility": 0.6000905632972717,
1730
+ "coverage": 1.0,
1731
+ "execute_generated": 1.0,
1732
+ "fallback_rate": 0.0,
1733
+ "hidden_chart_oracle_success": 1.0,
1734
+ "hidden_chart_oracle_utility": 2.0,
1735
+ "outcome_ptr": 0.0,
1736
+ "predicted_margin": 0.360946276685996,
1737
+ "proposal_oracle_success": 0.0,
1738
+ "proposal_oracle_utility": 0.21020369231700897,
1739
+ "selected_candidate_index": 4.0,
1740
+ "selected_success": 0.0,
1741
+ "selected_success_gain_over_base": 0.0,
1742
+ "selected_utility": 0.19976472854614258,
1743
+ "selected_utility_gain_over_base": -0.40032583475112915,
1744
+ "selector_regret": 0.010438963770866394,
1745
+ "success_selector_gap": 0.0,
1746
+ "success_support_gap": 1.0,
1747
+ "support_gap": 1.789796307682991,
1748
+ "tau": 0.288910690801741
1749
+ },
1750
+ "307068879/train1": {
1751
+ "base_success": 0.0,
1752
+ "base_utility": 0.6000905632972717,
1753
+ "coverage": 0.0,
1754
+ "execute_generated": 0.0,
1755
+ "fallback_rate": 1.0,
1756
+ "hidden_chart_oracle_success": 1.0,
1757
+ "hidden_chart_oracle_utility": 2.0,
1758
+ "outcome_ptr": 0.0,
1759
+ "predicted_margin": 0.26618249944242617,
1760
+ "proposal_oracle_success": 0.0,
1761
+ "proposal_oracle_utility": 0.21340669691562653,
1762
+ "selected_candidate_index": 4.0,
1763
+ "selected_success": 0.0,
1764
+ "selected_success_gain_over_base": 0.0,
1765
+ "selected_utility": 0.6000905632972717,
1766
+ "selected_utility_gain_over_base": 0.0,
1767
+ "selector_regret": 0.0,
1768
+ "success_selector_gap": 0.0,
1769
+ "success_support_gap": 1.0,
1770
+ "support_gap": 1.7865933030843735,
1771
+ "tau": 0.288910690801741
1772
+ },
1773
+ "307068879/train2": {
1774
+ "base_success": 0.0,
1775
+ "base_utility": 0.6000905632972717,
1776
+ "coverage": 1.0,
1777
+ "execute_generated": 1.0,
1778
+ "fallback_rate": 0.0,
1779
+ "hidden_chart_oracle_success": 1.0,
1780
+ "hidden_chart_oracle_utility": 2.0,
1781
+ "outcome_ptr": 1.0,
1782
+ "predicted_margin": 0.3071497785751483,
1783
+ "proposal_oracle_success": 1.0,
1784
+ "proposal_oracle_utility": 2.0,
1785
+ "selected_candidate_index": 6.0,
1786
+ "selected_success": 0.0,
1787
+ "selected_success_gain_over_base": 0.0,
1788
+ "selected_utility": 0.20327548682689667,
1789
+ "selected_utility_gain_over_base": -0.39681507647037506,
1790
+ "selector_regret": 1.7967245131731033,
1791
+ "success_selector_gap": 1.0,
1792
+ "success_support_gap": 0.0,
1793
+ "support_gap": 0.0,
1794
+ "tau": 0.288910690801741
1795
+ },
1796
+ "412102563/train0": {
1797
+ "base_success": 1.0,
1798
+ "base_utility": 2.0,
1799
+ "coverage": 1.0,
1800
+ "execute_generated": 1.0,
1801
+ "fallback_rate": 0.0,
1802
+ "hidden_chart_oracle_success": 1.0,
1803
+ "hidden_chart_oracle_utility": 2.0,
1804
+ "outcome_ptr": 0.0,
1805
+ "predicted_margin": 0.35814908299450166,
1806
+ "proposal_oracle_success": 1.0,
1807
+ "proposal_oracle_utility": 2.0,
1808
+ "selected_candidate_index": 2.0,
1809
+ "selected_success": 0.0,
1810
+ "selected_success_gain_over_base": -1.0,
1811
+ "selected_utility": 0.15394636988639832,
1812
+ "selected_utility_gain_over_base": -1.8460536301136017,
1813
+ "selector_regret": 1.8460536301136017,
1814
+ "success_selector_gap": 1.0,
1815
+ "success_support_gap": 0.0,
1816
+ "support_gap": 0.0,
1817
+ "tau": 0.288910690801741
1818
+ },
1819
+ "412102563/train1": {
1820
+ "base_success": 1.0,
1821
+ "base_utility": 2.0,
1822
+ "coverage": 1.0,
1823
+ "execute_generated": 1.0,
1824
+ "fallback_rate": 0.0,
1825
+ "hidden_chart_oracle_success": 1.0,
1826
+ "hidden_chart_oracle_utility": 2.0,
1827
+ "outcome_ptr": 0.0,
1828
+ "predicted_margin": 0.3186031235806744,
1829
+ "proposal_oracle_success": 1.0,
1830
+ "proposal_oracle_utility": 2.0,
1831
+ "selected_candidate_index": 2.0,
1832
+ "selected_success": 0.0,
1833
+ "selected_success_gain_over_base": -1.0,
1834
+ "selected_utility": 0.8125230669975281,
1835
+ "selected_utility_gain_over_base": -1.187476933002472,
1836
+ "selector_regret": 1.187476933002472,
1837
+ "success_selector_gap": 1.0,
1838
+ "success_support_gap": 0.0,
1839
+ "support_gap": 0.0,
1840
+ "tau": 0.288910690801741
1841
+ },
1842
+ "412102563/train2": {
1843
+ "base_success": 1.0,
1844
+ "base_utility": 2.0,
1845
+ "coverage": 1.0,
1846
+ "execute_generated": 1.0,
1847
+ "fallback_rate": 0.0,
1848
+ "hidden_chart_oracle_success": 1.0,
1849
+ "hidden_chart_oracle_utility": 2.0,
1850
+ "outcome_ptr": 0.0,
1851
+ "predicted_margin": 0.33548349915669784,
1852
+ "proposal_oracle_success": 1.0,
1853
+ "proposal_oracle_utility": 2.0,
1854
+ "selected_candidate_index": 5.0,
1855
+ "selected_success": 0.0,
1856
+ "selected_success_gain_over_base": -1.0,
1857
+ "selected_utility": 0.20452973246574402,
1858
+ "selected_utility_gain_over_base": -1.795470267534256,
1859
+ "selector_regret": 1.795470267534256,
1860
+ "success_selector_gap": 1.0,
1861
+ "success_support_gap": 0.0,
1862
+ "support_gap": 0.0,
1863
+ "tau": 0.288910690801741
1864
+ },
1865
+ "440577711/train0": {
1866
+ "base_success": 0.0,
1867
+ "base_utility": 0.036152780055999756,
1868
+ "coverage": 1.0,
1869
+ "execute_generated": 1.0,
1870
+ "fallback_rate": 0.0,
1871
+ "hidden_chart_oracle_success": 1.0,
1872
+ "hidden_chart_oracle_utility": 2.0,
1873
+ "outcome_ptr": 1.0,
1874
+ "predicted_margin": 0.373857640362683,
1875
+ "proposal_oracle_success": 1.0,
1876
+ "proposal_oracle_utility": 2.0,
1877
+ "selected_candidate_index": 1.0,
1878
+ "selected_success": 1.0,
1879
+ "selected_success_gain_over_base": 1.0,
1880
+ "selected_utility": 2.0,
1881
+ "selected_utility_gain_over_base": 1.9638472199440002,
1882
+ "selector_regret": 0.0,
1883
+ "success_selector_gap": 0.0,
1884
+ "success_support_gap": 0.0,
1885
+ "support_gap": 0.0,
1886
+ "tau": 0.288910690801741
1887
+ },
1888
+ "440577711/train1": {
1889
+ "base_success": 0.0,
1890
+ "base_utility": 0.036152780055999756,
1891
+ "coverage": 1.0,
1892
+ "execute_generated": 1.0,
1893
+ "fallback_rate": 0.0,
1894
+ "hidden_chart_oracle_success": 1.0,
1895
+ "hidden_chart_oracle_utility": 2.0,
1896
+ "outcome_ptr": 1.0,
1897
+ "predicted_margin": 0.4239981896253331,
1898
+ "proposal_oracle_success": 1.0,
1899
+ "proposal_oracle_utility": 2.0,
1900
+ "selected_candidate_index": 3.0,
1901
+ "selected_success": 1.0,
1902
+ "selected_success_gain_over_base": 1.0,
1903
+ "selected_utility": 2.0,
1904
+ "selected_utility_gain_over_base": 1.9638472199440002,
1905
+ "selector_regret": 0.0,
1906
+ "success_selector_gap": 0.0,
1907
+ "success_support_gap": 0.0,
1908
+ "support_gap": 0.0,
1909
+ "tau": 0.288910690801741
1910
+ },
1911
+ "440577711/train2": {
1912
+ "base_success": 0.0,
1913
+ "base_utility": 0.036152780055999756,
1914
+ "coverage": 1.0,
1915
+ "execute_generated": 1.0,
1916
+ "fallback_rate": 0.0,
1917
+ "hidden_chart_oracle_success": 1.0,
1918
+ "hidden_chart_oracle_utility": 2.0,
1919
+ "outcome_ptr": 1.0,
1920
+ "predicted_margin": 0.43744374451697843,
1921
+ "proposal_oracle_success": 1.0,
1922
+ "proposal_oracle_utility": 2.0,
1923
+ "selected_candidate_index": 1.0,
1924
+ "selected_success": 1.0,
1925
+ "selected_success_gain_over_base": 1.0,
1926
+ "selected_utility": 2.0,
1927
+ "selected_utility_gain_over_base": 1.9638472199440002,
1928
+ "selector_regret": 0.0,
1929
+ "success_selector_gap": 0.0,
1930
+ "success_support_gap": 0.0,
1931
+ "support_gap": 0.0,
1932
+ "tau": 0.288910690801741
1933
+ },
1934
+ "441046872/train0": {
1935
+ "base_success": 0.0,
1936
+ "base_utility": 0.6999551653862,
1937
+ "coverage": 1.0,
1938
+ "execute_generated": 1.0,
1939
+ "fallback_rate": 0.0,
1940
+ "hidden_chart_oracle_success": 1.0,
1941
+ "hidden_chart_oracle_utility": 2.0,
1942
+ "outcome_ptr": 0.0,
1943
+ "predicted_margin": 0.36230088488387413,
1944
+ "proposal_oracle_success": 0.0,
1945
+ "proposal_oracle_utility": 0.5237247943878174,
1946
+ "selected_candidate_index": 1.0,
1947
+ "selected_success": 0.0,
1948
+ "selected_success_gain_over_base": 0.0,
1949
+ "selected_utility": 0.21159441769123077,
1950
+ "selected_utility_gain_over_base": -0.4883607476949692,
1951
+ "selector_regret": 0.3121303766965866,
1952
+ "success_selector_gap": 0.0,
1953
+ "success_support_gap": 1.0,
1954
+ "support_gap": 1.4762752056121826,
1955
+ "tau": 0.288910690801741
1956
+ },
1957
+ "441046872/train1": {
1958
+ "base_success": 0.0,
1959
+ "base_utility": 0.6999551653862,
1960
+ "coverage": 0.0,
1961
+ "execute_generated": 0.0,
1962
+ "fallback_rate": 1.0,
1963
+ "hidden_chart_oracle_success": 1.0,
1964
+ "hidden_chart_oracle_utility": 2.0,
1965
+ "outcome_ptr": 1.0,
1966
+ "predicted_margin": 0.26771412761361,
1967
+ "proposal_oracle_success": 1.0,
1968
+ "proposal_oracle_utility": 2.0,
1969
+ "selected_candidate_index": 2.0,
1970
+ "selected_success": 0.0,
1971
+ "selected_success_gain_over_base": 0.0,
1972
+ "selected_utility": 0.6999551653862,
1973
+ "selected_utility_gain_over_base": 0.0,
1974
+ "selector_regret": 1.3000448346138,
1975
+ "success_selector_gap": 1.0,
1976
+ "success_support_gap": 0.0,
1977
+ "support_gap": 0.0,
1978
+ "tau": 0.288910690801741
1979
+ },
1980
+ "441046872/train2": {
1981
+ "base_success": 0.0,
1982
+ "base_utility": 0.6999551653862,
1983
+ "coverage": 1.0,
1984
+ "execute_generated": 1.0,
1985
+ "fallback_rate": 0.0,
1986
+ "hidden_chart_oracle_success": 1.0,
1987
+ "hidden_chart_oracle_utility": 2.0,
1988
+ "outcome_ptr": 1.0,
1989
+ "predicted_margin": 0.3190949511854425,
1990
+ "proposal_oracle_success": 1.0,
1991
+ "proposal_oracle_utility": 2.0,
1992
+ "selected_candidate_index": 1.0,
1993
+ "selected_success": 0.0,
1994
+ "selected_success_gain_over_base": 0.0,
1995
+ "selected_utility": 0.5507636666297913,
1996
+ "selected_utility_gain_over_base": -0.1491914987564087,
1997
+ "selector_regret": 1.4492363333702087,
1998
+ "success_selector_gap": 1.0,
1999
+ "success_support_gap": 0.0,
2000
+ "support_gap": 0.0,
2001
+ "tau": 0.288910690801741
2002
+ },
2003
+ "509430392/train0": {
2004
+ "base_success": 0.0,
2005
+ "base_utility": 2.2351741790771484e-07,
2006
+ "coverage": 1.0,
2007
+ "execute_generated": 1.0,
2008
+ "fallback_rate": 0.0,
2009
+ "hidden_chart_oracle_success": 1.0,
2010
+ "hidden_chart_oracle_utility": 2.0,
2011
+ "outcome_ptr": 1.0,
2012
+ "predicted_margin": 0.32410094565879394,
2013
+ "proposal_oracle_success": 1.0,
2014
+ "proposal_oracle_utility": 2.0,
2015
+ "selected_candidate_index": 1.0,
2016
+ "selected_success": 0.0,
2017
+ "selected_success_gain_over_base": 0.0,
2018
+ "selected_utility": 1.1026859283447266e-06,
2019
+ "selected_utility_gain_over_base": 8.791685104370117e-07,
2020
+ "selector_regret": 1.9999988973140717,
2021
+ "success_selector_gap": 1.0,
2022
+ "success_support_gap": 0.0,
2023
+ "support_gap": 0.0,
2024
+ "tau": 0.288910690801741
2025
+ },
2026
+ "509430392/train1": {
2027
+ "base_success": 0.0,
2028
+ "base_utility": 2.2351741790771484e-07,
2029
+ "coverage": 1.0,
2030
+ "execute_generated": 1.0,
2031
+ "fallback_rate": 0.0,
2032
+ "hidden_chart_oracle_success": 1.0,
2033
+ "hidden_chart_oracle_utility": 2.0,
2034
+ "outcome_ptr": 1.0,
2035
+ "predicted_margin": 0.31748984216726817,
2036
+ "proposal_oracle_success": 0.0,
2037
+ "proposal_oracle_utility": 2.028048038482666e-05,
2038
+ "selected_candidate_index": 1.0,
2039
+ "selected_success": 0.0,
2040
+ "selected_success_gain_over_base": 0.0,
2041
+ "selected_utility": 1.5348196029663086e-06,
2042
+ "selected_utility_gain_over_base": 1.3113021850585938e-06,
2043
+ "selector_regret": 1.874566078186035e-05,
2044
+ "success_selector_gap": 0.0,
2045
+ "success_support_gap": 1.0,
2046
+ "support_gap": 1.9999797195196152,
2047
+ "tau": 0.288910690801741
2048
+ },
2049
+ "509430392/train2": {
2050
+ "base_success": 0.0,
2051
+ "base_utility": 2.2351741790771484e-07,
2052
+ "coverage": 0.0,
2053
+ "execute_generated": 0.0,
2054
+ "fallback_rate": 1.0,
2055
+ "hidden_chart_oracle_success": 1.0,
2056
+ "hidden_chart_oracle_utility": 2.0,
2057
+ "outcome_ptr": 1.0,
2058
+ "predicted_margin": 0.2809423707634872,
2059
+ "proposal_oracle_success": 0.0,
2060
+ "proposal_oracle_utility": 3.084540367126465e-06,
2061
+ "selected_candidate_index": 2.0,
2062
+ "selected_success": 0.0,
2063
+ "selected_success_gain_over_base": 0.0,
2064
+ "selected_utility": 2.2351741790771484e-07,
2065
+ "selected_utility_gain_over_base": 0.0,
2066
+ "selector_regret": 2.86102294921875e-06,
2067
+ "success_selector_gap": 0.0,
2068
+ "success_support_gap": 1.0,
2069
+ "support_gap": 1.9999969154596329,
2070
+ "tau": 0.288910690801741
2071
+ },
2072
+ "585904130/train0": {
2073
+ "base_success": 0.0,
2074
+ "base_utility": 0.18171554803848267,
2075
+ "coverage": 0.0,
2076
+ "execute_generated": 0.0,
2077
+ "fallback_rate": 1.0,
2078
+ "hidden_chart_oracle_success": 1.0,
2079
+ "hidden_chart_oracle_utility": 2.0,
2080
+ "outcome_ptr": 1.0,
2081
+ "predicted_margin": 0.24217696385677426,
2082
+ "proposal_oracle_success": 1.0,
2083
+ "proposal_oracle_utility": 2.0,
2084
+ "selected_candidate_index": 7.0,
2085
+ "selected_success": 0.0,
2086
+ "selected_success_gain_over_base": 0.0,
2087
+ "selected_utility": 0.18171554803848267,
2088
+ "selected_utility_gain_over_base": 0.0,
2089
+ "selector_regret": 1.8182844519615173,
2090
+ "success_selector_gap": 1.0,
2091
+ "success_support_gap": 0.0,
2092
+ "support_gap": 0.0,
2093
+ "tau": 0.288910690801741
2094
+ },
2095
+ "585904130/train1": {
2096
+ "base_success": 0.0,
2097
+ "base_utility": 0.18171554803848267,
2098
+ "coverage": 0.0,
2099
+ "execute_generated": 0.0,
2100
+ "fallback_rate": 1.0,
2101
+ "hidden_chart_oracle_success": 1.0,
2102
+ "hidden_chart_oracle_utility": 2.0,
2103
+ "outcome_ptr": 1.0,
2104
+ "predicted_margin": 0.23626096133300956,
2105
+ "proposal_oracle_success": 1.0,
2106
+ "proposal_oracle_utility": 2.0,
2107
+ "selected_candidate_index": 4.0,
2108
+ "selected_success": 0.0,
2109
+ "selected_success_gain_over_base": 0.0,
2110
+ "selected_utility": 0.18171554803848267,
2111
+ "selected_utility_gain_over_base": 0.0,
2112
+ "selector_regret": 1.8182844519615173,
2113
+ "success_selector_gap": 1.0,
2114
+ "success_support_gap": 0.0,
2115
+ "support_gap": 0.0,
2116
+ "tau": 0.288910690801741
2117
+ },
2118
+ "585904130/train2": {
2119
+ "base_success": 0.0,
2120
+ "base_utility": 0.18171554803848267,
2121
+ "coverage": 1.0,
2122
+ "execute_generated": 1.0,
2123
+ "fallback_rate": 0.0,
2124
+ "hidden_chart_oracle_success": 1.0,
2125
+ "hidden_chart_oracle_utility": 2.0,
2126
+ "outcome_ptr": 1.0,
2127
+ "predicted_margin": 0.29205996264326795,
2128
+ "proposal_oracle_success": 1.0,
2129
+ "proposal_oracle_utility": 2.0,
2130
+ "selected_candidate_index": 0.0,
2131
+ "selected_success": 1.0,
2132
+ "selected_success_gain_over_base": 1.0,
2133
+ "selected_utility": 2.0,
2134
+ "selected_utility_gain_over_base": 1.8182844519615173,
2135
+ "selector_regret": 0.0,
2136
+ "success_selector_gap": 0.0,
2137
+ "success_support_gap": 0.0,
2138
+ "support_gap": 0.0,
2139
+ "tau": 0.288910690801741
2140
+ },
2141
+ "608016290/train0": {
2142
+ "base_success": 0.0,
2143
+ "base_utility": 0.725701630115509,
2144
+ "coverage": 1.0,
2145
+ "execute_generated": 1.0,
2146
+ "fallback_rate": 0.0,
2147
+ "hidden_chart_oracle_success": 1.0,
2148
+ "hidden_chart_oracle_utility": 2.0,
2149
+ "outcome_ptr": 1.0,
2150
+ "predicted_margin": 0.36311287488276633,
2151
+ "proposal_oracle_success": 1.0,
2152
+ "proposal_oracle_utility": 2.0,
2153
+ "selected_candidate_index": 4.0,
2154
+ "selected_success": 1.0,
2155
+ "selected_success_gain_over_base": 1.0,
2156
+ "selected_utility": 2.0,
2157
+ "selected_utility_gain_over_base": 1.274298369884491,
2158
+ "selector_regret": 0.0,
2159
+ "success_selector_gap": 0.0,
2160
+ "success_support_gap": 0.0,
2161
+ "support_gap": 0.0,
2162
+ "tau": 0.288910690801741
2163
+ },
2164
+ "608016290/train1": {
2165
+ "base_success": 0.0,
2166
+ "base_utility": 0.725701630115509,
2167
+ "coverage": 1.0,
2168
+ "execute_generated": 1.0,
2169
+ "fallback_rate": 0.0,
2170
+ "hidden_chart_oracle_success": 1.0,
2171
+ "hidden_chart_oracle_utility": 2.0,
2172
+ "outcome_ptr": 1.0,
2173
+ "predicted_margin": 0.33152683041541503,
2174
+ "proposal_oracle_success": 1.0,
2175
+ "proposal_oracle_utility": 2.0,
2176
+ "selected_candidate_index": 1.0,
2177
+ "selected_success": 1.0,
2178
+ "selected_success_gain_over_base": 1.0,
2179
+ "selected_utility": 2.0,
2180
+ "selected_utility_gain_over_base": 1.274298369884491,
2181
+ "selector_regret": 0.0,
2182
+ "success_selector_gap": 0.0,
2183
+ "success_support_gap": 0.0,
2184
+ "support_gap": 0.0,
2185
+ "tau": 0.288910690801741
2186
+ },
2187
+ "608016290/train2": {
2188
+ "base_success": 0.0,
2189
+ "base_utility": 0.725701630115509,
2190
+ "coverage": 1.0,
2191
+ "execute_generated": 1.0,
2192
+ "fallback_rate": 0.0,
2193
+ "hidden_chart_oracle_success": 1.0,
2194
+ "hidden_chart_oracle_utility": 2.0,
2195
+ "outcome_ptr": 1.0,
2196
+ "predicted_margin": 0.3155860097253581,
2197
+ "proposal_oracle_success": 1.0,
2198
+ "proposal_oracle_utility": 2.0,
2199
+ "selected_candidate_index": 3.0,
2200
+ "selected_success": 0.0,
2201
+ "selected_success_gain_over_base": 0.0,
2202
+ "selected_utility": 0.21869932115077972,
2203
+ "selected_utility_gain_over_base": -0.5070023089647293,
2204
+ "selector_regret": 1.7813006788492203,
2205
+ "success_selector_gap": 1.0,
2206
+ "success_support_gap": 0.0,
2207
+ "support_gap": 0.0,
2208
+ "tau": 0.288910690801741
2209
+ },
2210
+ "626486813/train0": {
2211
+ "base_success": 0.0,
2212
+ "base_utility": 0.00509834272088483,
2213
+ "coverage": 1.0,
2214
+ "execute_generated": 1.0,
2215
+ "fallback_rate": 0.0,
2216
+ "hidden_chart_oracle_success": 0.0,
2217
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2218
+ "outcome_ptr": 1.0,
2219
+ "predicted_margin": 0.3010244399554821,
2220
+ "proposal_oracle_success": 0.0,
2221
+ "proposal_oracle_utility": 0.041232248302549124,
2222
+ "selected_candidate_index": 6.0,
2223
+ "selected_success": 0.0,
2224
+ "selected_success_gain_over_base": 0.0,
2225
+ "selected_utility": 0.007493555312976241,
2226
+ "selected_utility_gain_over_base": 0.0023952125920914114,
2227
+ "selector_regret": 0.03373869298957288,
2228
+ "success_selector_gap": 0.0,
2229
+ "success_support_gap": 0.0,
2230
+ "support_gap": 0.12435013288632035,
2231
+ "tau": 0.288910690801741
2232
+ },
2233
+ "626486813/train1": {
2234
+ "base_success": 0.0,
2235
+ "base_utility": 0.00509834272088483,
2236
+ "coverage": 1.0,
2237
+ "execute_generated": 1.0,
2238
+ "fallback_rate": 0.0,
2239
+ "hidden_chart_oracle_success": 0.0,
2240
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2241
+ "outcome_ptr": 1.0,
2242
+ "predicted_margin": 0.3307195254789452,
2243
+ "proposal_oracle_success": 0.0,
2244
+ "proposal_oracle_utility": 0.02825745940208435,
2245
+ "selected_candidate_index": 4.5,
2246
+ "selected_success": 0.0,
2247
+ "selected_success_gain_over_base": 0.0,
2248
+ "selected_utility": 0.02127585862763226,
2249
+ "selected_utility_gain_over_base": 0.01617751590674743,
2250
+ "selector_regret": 0.00698160077445209,
2251
+ "success_selector_gap": 0.0,
2252
+ "success_support_gap": 0.0,
2253
+ "support_gap": 0.13732492178678513,
2254
+ "tau": 0.288910690801741
2255
+ },
2256
+ "626486813/train2": {
2257
+ "base_success": 0.0,
2258
+ "base_utility": 0.00509834272088483,
2259
+ "coverage": 1.0,
2260
+ "execute_generated": 1.0,
2261
+ "fallback_rate": 0.0,
2262
+ "hidden_chart_oracle_success": 0.0,
2263
+ "hidden_chart_oracle_utility": 0.16558238118886948,
2264
+ "outcome_ptr": 1.0,
2265
+ "predicted_margin": 0.33842261077881797,
2266
+ "proposal_oracle_success": 0.0,
2267
+ "proposal_oracle_utility": 0.03251718543469906,
2268
+ "selected_candidate_index": 1.0,
2269
+ "selected_success": 0.0,
2270
+ "selected_success_gain_over_base": 0.0,
2271
+ "selected_utility": 0.020153234712779522,
2272
+ "selected_utility_gain_over_base": 0.015054891991894692,
2273
+ "selector_regret": 0.012363950721919537,
2274
+ "success_selector_gap": 0.0,
2275
+ "success_support_gap": 0.0,
2276
+ "support_gap": 0.13306519575417042,
2277
+ "tau": 0.288910690801741
2278
+ },
2279
+ "636680137/train0": {
2280
+ "base_success": 0.0,
2281
+ "base_utility": 0.08313504606485367,
2282
+ "coverage": 1.0,
2283
+ "execute_generated": 1.0,
2284
+ "fallback_rate": 0.0,
2285
+ "hidden_chart_oracle_success": 0.0,
2286
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2287
+ "outcome_ptr": 1.0,
2288
+ "predicted_margin": 0.3370754671183242,
2289
+ "proposal_oracle_success": 0.0,
2290
+ "proposal_oracle_utility": 0.08482346683740616,
2291
+ "selected_candidate_index": 1.0,
2292
+ "selected_success": 0.0,
2293
+ "selected_success_gain_over_base": 0.0,
2294
+ "selected_utility": 0.08482346683740616,
2295
+ "selected_utility_gain_over_base": 0.0016884207725524902,
2296
+ "selector_regret": 0.0,
2297
+ "success_selector_gap": 0.0,
2298
+ "success_support_gap": 0.0,
2299
+ "support_gap": 0.08574085682630539,
2300
+ "tau": 0.288910690801741
2301
+ },
2302
+ "636680137/train1": {
2303
+ "base_success": 0.0,
2304
+ "base_utility": 0.08313504606485367,
2305
+ "coverage": 0.0,
2306
+ "execute_generated": 0.0,
2307
+ "fallback_rate": 1.0,
2308
+ "hidden_chart_oracle_success": 0.0,
2309
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2310
+ "outcome_ptr": 1.0,
2311
+ "predicted_margin": 0.2691169427030669,
2312
+ "proposal_oracle_success": 0.0,
2313
+ "proposal_oracle_utility": 0.08613479137420654,
2314
+ "selected_candidate_index": 1.0,
2315
+ "selected_success": 0.0,
2316
+ "selected_success_gain_over_base": 0.0,
2317
+ "selected_utility": 0.08313504606485367,
2318
+ "selected_utility_gain_over_base": 0.0,
2319
+ "selector_regret": 0.0029997453093528748,
2320
+ "success_selector_gap": 0.0,
2321
+ "success_support_gap": 0.0,
2322
+ "support_gap": 0.084429532289505,
2323
+ "tau": 0.288910690801741
2324
+ },
2325
+ "636680137/train2": {
2326
+ "base_success": 0.0,
2327
+ "base_utility": 0.08313504606485367,
2328
+ "coverage": 1.0,
2329
+ "execute_generated": 1.0,
2330
+ "fallback_rate": 0.0,
2331
+ "hidden_chart_oracle_success": 0.0,
2332
+ "hidden_chart_oracle_utility": 0.17056432366371155,
2333
+ "outcome_ptr": 0.0,
2334
+ "predicted_margin": 0.311291997921847,
2335
+ "proposal_oracle_success": 0.0,
2336
+ "proposal_oracle_utility": 0.0808148980140686,
2337
+ "selected_candidate_index": 7.0,
2338
+ "selected_success": 0.0,
2339
+ "selected_success_gain_over_base": 0.0,
2340
+ "selected_utility": 0.04636569693684578,
2341
+ "selected_utility_gain_over_base": -0.03676934912800789,
2342
+ "selector_regret": 0.034449201077222824,
2343
+ "success_selector_gap": 0.0,
2344
+ "success_support_gap": 0.0,
2345
+ "support_gap": 0.08974942564964294,
2346
+ "tau": 0.288910690801741
2347
+ },
2348
+ "655884605/train0": {
2349
+ "base_success": 0.0,
2350
+ "base_utility": 0.16300667822360992,
2351
+ "coverage": 1.0,
2352
+ "execute_generated": 1.0,
2353
+ "fallback_rate": 0.0,
2354
+ "hidden_chart_oracle_success": 1.0,
2355
+ "hidden_chart_oracle_utility": 2.0,
2356
+ "outcome_ptr": 0.0,
2357
+ "predicted_margin": 0.31969823977925854,
2358
+ "proposal_oracle_success": 0.0,
2359
+ "proposal_oracle_utility": 0.15542201697826385,
2360
+ "selected_candidate_index": 5.0,
2361
+ "selected_success": 0.0,
2362
+ "selected_success_gain_over_base": 0.0,
2363
+ "selected_utility": 0.15542201697826385,
2364
+ "selected_utility_gain_over_base": -0.007584661245346069,
2365
+ "selector_regret": 0.0,
2366
+ "success_selector_gap": 0.0,
2367
+ "success_support_gap": 1.0,
2368
+ "support_gap": 1.8445779830217361,
2369
+ "tau": 0.288910690801741
2370
+ },
2371
+ "655884605/train1": {
2372
+ "base_success": 0.0,
2373
+ "base_utility": 0.16300667822360992,
2374
+ "coverage": 1.0,
2375
+ "execute_generated": 1.0,
2376
+ "fallback_rate": 0.0,
2377
+ "hidden_chart_oracle_success": 1.0,
2378
+ "hidden_chart_oracle_utility": 2.0,
2379
+ "outcome_ptr": 1.0,
2380
+ "predicted_margin": 0.30504445185011525,
2381
+ "proposal_oracle_success": 0.0,
2382
+ "proposal_oracle_utility": 0.8609292507171631,
2383
+ "selected_candidate_index": 5.0,
2384
+ "selected_success": 0.0,
2385
+ "selected_success_gain_over_base": 0.0,
2386
+ "selected_utility": 0.10443197190761566,
2387
+ "selected_utility_gain_over_base": -0.05857470631599426,
2388
+ "selector_regret": 0.7564972788095474,
2389
+ "success_selector_gap": 0.0,
2390
+ "success_support_gap": 1.0,
2391
+ "support_gap": 1.139070749282837,
2392
+ "tau": 0.288910690801741
2393
+ },
2394
+ "655884605/train2": {
2395
+ "base_success": 0.0,
2396
+ "base_utility": 0.16300667822360992,
2397
+ "coverage": 1.0,
2398
+ "execute_generated": 1.0,
2399
+ "fallback_rate": 0.0,
2400
+ "hidden_chart_oracle_success": 1.0,
2401
+ "hidden_chart_oracle_utility": 2.0,
2402
+ "outcome_ptr": 1.0,
2403
+ "predicted_margin": 0.31871607418195164,
2404
+ "proposal_oracle_success": 0.0,
2405
+ "proposal_oracle_utility": 0.22648370265960693,
2406
+ "selected_candidate_index": 2.0,
2407
+ "selected_success": 0.0,
2408
+ "selected_success_gain_over_base": 0.0,
2409
+ "selected_utility": 0.07072655856609344,
2410
+ "selected_utility_gain_over_base": -0.09228011965751648,
2411
+ "selector_regret": 0.1557571440935135,
2412
+ "success_selector_gap": 0.0,
2413
+ "success_support_gap": 1.0,
2414
+ "support_gap": 1.773516297340393,
2415
+ "tau": 0.288910690801741
2416
+ },
2417
+ "672328262/train0": {
2418
+ "base_success": 1.0,
2419
+ "base_utility": 2.0,
2420
+ "coverage": 0.0,
2421
+ "execute_generated": 0.0,
2422
+ "fallback_rate": 1.0,
2423
+ "hidden_chart_oracle_success": 1.0,
2424
+ "hidden_chart_oracle_utility": 2.0,
2425
+ "outcome_ptr": 0.0,
2426
+ "predicted_margin": 0.28545046297241927,
2427
+ "proposal_oracle_success": 1.0,
2428
+ "proposal_oracle_utility": 2.0,
2429
+ "selected_candidate_index": 6.0,
2430
+ "selected_success": 1.0,
2431
+ "selected_success_gain_over_base": 0.0,
2432
+ "selected_utility": 2.0,
2433
+ "selected_utility_gain_over_base": 0.0,
2434
+ "selector_regret": 0.0,
2435
+ "success_selector_gap": 0.0,
2436
+ "success_support_gap": 0.0,
2437
+ "support_gap": 0.0,
2438
+ "tau": 0.288910690801741
2439
+ },
2440
+ "672328262/train1": {
2441
+ "base_success": 1.0,
2442
+ "base_utility": 2.0,
2443
+ "coverage": 1.0,
2444
+ "execute_generated": 1.0,
2445
+ "fallback_rate": 0.0,
2446
+ "hidden_chart_oracle_success": 1.0,
2447
+ "hidden_chart_oracle_utility": 2.0,
2448
+ "outcome_ptr": 0.0,
2449
+ "predicted_margin": 0.3101015251021578,
2450
+ "proposal_oracle_success": 1.0,
2451
+ "proposal_oracle_utility": 2.0,
2452
+ "selected_candidate_index": 1.0,
2453
+ "selected_success": 1.0,
2454
+ "selected_success_gain_over_base": 0.0,
2455
+ "selected_utility": 2.0,
2456
+ "selected_utility_gain_over_base": 0.0,
2457
+ "selector_regret": 0.0,
2458
+ "success_selector_gap": 0.0,
2459
+ "success_support_gap": 0.0,
2460
+ "support_gap": 0.0,
2461
+ "tau": 0.288910690801741
2462
+ },
2463
+ "672328262/train2": {
2464
+ "base_success": 1.0,
2465
+ "base_utility": 2.0,
2466
+ "coverage": 1.0,
2467
+ "execute_generated": 1.0,
2468
+ "fallback_rate": 0.0,
2469
+ "hidden_chart_oracle_success": 1.0,
2470
+ "hidden_chart_oracle_utility": 2.0,
2471
+ "outcome_ptr": 0.0,
2472
+ "predicted_margin": 0.29576444314052996,
2473
+ "proposal_oracle_success": 1.0,
2474
+ "proposal_oracle_utility": 2.0,
2475
+ "selected_candidate_index": 3.0,
2476
+ "selected_success": 0.0,
2477
+ "selected_success_gain_over_base": -1.0,
2478
+ "selected_utility": 0.08623594045639038,
2479
+ "selected_utility_gain_over_base": -1.9137640595436096,
2480
+ "selector_regret": 1.9137640595436096,
2481
+ "success_selector_gap": 1.0,
2482
+ "success_support_gap": 0.0,
2483
+ "support_gap": 0.0,
2484
+ "tau": 0.288910690801741
2485
+ },
2486
+ "701667486/train0": {
2487
+ "base_success": 1.0,
2488
+ "base_utility": 2.0,
2489
+ "coverage": 1.0,
2490
+ "execute_generated": 1.0,
2491
+ "fallback_rate": 0.0,
2492
+ "hidden_chart_oracle_success": 1.0,
2493
+ "hidden_chart_oracle_utility": 2.0,
2494
+ "outcome_ptr": 0.0,
2495
+ "predicted_margin": 0.3235847894999446,
2496
+ "proposal_oracle_success": 1.0,
2497
+ "proposal_oracle_utility": 2.0,
2498
+ "selected_candidate_index": 3.0,
2499
+ "selected_success": 1.0,
2500
+ "selected_success_gain_over_base": 0.0,
2501
+ "selected_utility": 2.0,
2502
+ "selected_utility_gain_over_base": 0.0,
2503
+ "selector_regret": 0.0,
2504
+ "success_selector_gap": 0.0,
2505
+ "success_support_gap": 0.0,
2506
+ "support_gap": 0.0,
2507
+ "tau": 0.288910690801741
2508
+ },
2509
+ "701667486/train1": {
2510
+ "base_success": 1.0,
2511
+ "base_utility": 2.0,
2512
+ "coverage": 1.0,
2513
+ "execute_generated": 1.0,
2514
+ "fallback_rate": 0.0,
2515
+ "hidden_chart_oracle_success": 1.0,
2516
+ "hidden_chart_oracle_utility": 2.0,
2517
+ "outcome_ptr": 0.0,
2518
+ "predicted_margin": 0.3672899579742953,
2519
+ "proposal_oracle_success": 1.0,
2520
+ "proposal_oracle_utility": 2.0,
2521
+ "selected_candidate_index": 4.0,
2522
+ "selected_success": 1.0,
2523
+ "selected_success_gain_over_base": 0.0,
2524
+ "selected_utility": 2.0,
2525
+ "selected_utility_gain_over_base": 0.0,
2526
+ "selector_regret": 0.0,
2527
+ "success_selector_gap": 0.0,
2528
+ "success_support_gap": 0.0,
2529
+ "support_gap": 0.0,
2530
+ "tau": 0.288910690801741
2531
+ },
2532
+ "701667486/train2": {
2533
+ "base_success": 1.0,
2534
+ "base_utility": 2.0,
2535
+ "coverage": 1.0,
2536
+ "execute_generated": 1.0,
2537
+ "fallback_rate": 0.0,
2538
+ "hidden_chart_oracle_success": 1.0,
2539
+ "hidden_chart_oracle_utility": 2.0,
2540
+ "outcome_ptr": 0.0,
2541
+ "predicted_margin": 0.3705436174062195,
2542
+ "proposal_oracle_success": 1.0,
2543
+ "proposal_oracle_utility": 2.0,
2544
+ "selected_candidate_index": 3.0,
2545
+ "selected_success": 0.0,
2546
+ "selected_success_gain_over_base": -1.0,
2547
+ "selected_utility": 0.20135055482387543,
2548
+ "selected_utility_gain_over_base": -1.7986494451761246,
2549
+ "selector_regret": 1.7986494451761246,
2550
+ "success_selector_gap": 1.0,
2551
+ "success_support_gap": 0.0,
2552
+ "support_gap": 0.0,
2553
+ "tau": 0.288910690801741
2554
+ },
2555
+ "747260437/train0": {
2556
+ "base_success": 1.0,
2557
+ "base_utility": 2.0,
2558
+ "coverage": 0.0,
2559
+ "execute_generated": 0.0,
2560
+ "fallback_rate": 1.0,
2561
+ "hidden_chart_oracle_success": 0.0,
2562
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2563
+ "outcome_ptr": 0.0,
2564
+ "predicted_margin": 0.2642116780504518,
2565
+ "proposal_oracle_success": 1.0,
2566
+ "proposal_oracle_utility": 2.0,
2567
+ "selected_candidate_index": 0.0,
2568
+ "selected_success": 1.0,
2569
+ "selected_success_gain_over_base": 0.0,
2570
+ "selected_utility": 2.0,
2571
+ "selected_utility_gain_over_base": 0.0,
2572
+ "selector_regret": 0.0,
2573
+ "success_selector_gap": 0.0,
2574
+ "success_support_gap": 0.0,
2575
+ "support_gap": 0.0,
2576
+ "tau": 0.288910690801741
2577
+ },
2578
+ "747260437/train1": {
2579
+ "base_success": 1.0,
2580
+ "base_utility": 2.0,
2581
+ "coverage": 0.0,
2582
+ "execute_generated": 0.0,
2583
+ "fallback_rate": 1.0,
2584
+ "hidden_chart_oracle_success": 0.0,
2585
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2586
+ "outcome_ptr": 0.0,
2587
+ "predicted_margin": 0.28462172278681314,
2588
+ "proposal_oracle_success": 0.0,
2589
+ "proposal_oracle_utility": 0.05061175301671028,
2590
+ "selected_candidate_index": 6.0,
2591
+ "selected_success": 1.0,
2592
+ "selected_success_gain_over_base": 0.0,
2593
+ "selected_utility": 2.0,
2594
+ "selected_utility_gain_over_base": 0.0,
2595
+ "selector_regret": 0.0,
2596
+ "success_selector_gap": 0.0,
2597
+ "success_support_gap": 0.0,
2598
+ "support_gap": 0.22805149480700493,
2599
+ "tau": 0.288910690801741
2600
+ },
2601
+ "747260437/train2": {
2602
+ "base_success": 1.0,
2603
+ "base_utility": 2.0,
2604
+ "coverage": 1.0,
2605
+ "execute_generated": 1.0,
2606
+ "fallback_rate": 0.0,
2607
+ "hidden_chart_oracle_success": 0.0,
2608
+ "hidden_chart_oracle_utility": 0.2786632478237152,
2609
+ "outcome_ptr": 0.0,
2610
+ "predicted_margin": 0.2893617063491729,
2611
+ "proposal_oracle_success": 0.0,
2612
+ "proposal_oracle_utility": 0.11570342630147934,
2613
+ "selected_candidate_index": 4.0,
2614
+ "selected_success": 0.0,
2615
+ "selected_success_gain_over_base": -1.0,
2616
+ "selected_utility": 0.049731094390153885,
2617
+ "selected_utility_gain_over_base": -1.9502689056098461,
2618
+ "selector_regret": 0.06597233191132545,
2619
+ "success_selector_gap": 0.0,
2620
+ "success_support_gap": 0.0,
2621
+ "support_gap": 0.16295982152223587,
2622
+ "tau": 0.288910690801741
2623
+ },
2624
+ "750448296/train0": {
2625
+ "base_success": 0.0,
2626
+ "base_utility": 0.11759314934412639,
2627
+ "coverage": 0.3333333333333333,
2628
+ "execute_generated": 0.3333333333333333,
2629
+ "fallback_rate": 0.6666666666666666,
2630
+ "hidden_chart_oracle_success": 0.6666666666666666,
2631
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2632
+ "outcome_ptr": 0.6666666666666666,
2633
+ "predicted_margin": 0.2639246365033728,
2634
+ "proposal_oracle_success": 0.3333333333333333,
2635
+ "proposal_oracle_utility": 0.7487806566059589,
2636
+ "selected_candidate_index": 1.0,
2637
+ "selected_success": 0.0,
2638
+ "selected_success_gain_over_base": 0.0,
2639
+ "selected_utility": 0.10066552336017291,
2640
+ "selected_utility_gain_over_base": -0.016927625983953476,
2641
+ "selector_regret": 0.6481151332457861,
2642
+ "success_selector_gap": 0.3333333333333333,
2643
+ "success_support_gap": 0.3333333333333333,
2644
+ "support_gap": 0.6592158389588197,
2645
+ "tau": 0.288910690801741
2646
+ },
2647
+ "750448296/train1": {
2648
+ "base_success": 0.0,
2649
+ "base_utility": 0.11759314934412639,
2650
+ "coverage": 0.3333333333333333,
2651
+ "execute_generated": 0.3333333333333333,
2652
+ "fallback_rate": 0.6666666666666666,
2653
+ "hidden_chart_oracle_success": 0.6666666666666666,
2654
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2655
+ "outcome_ptr": 1.0,
2656
+ "predicted_margin": 0.28191903992498957,
2657
+ "proposal_oracle_success": 0.6666666666666666,
2658
+ "proposal_oracle_utility": 1.3738718951741855,
2659
+ "selected_candidate_index": 3.0,
2660
+ "selected_success": 0.0,
2661
+ "selected_success_gain_over_base": 0.0,
2662
+ "selected_utility": 0.11092466861009598,
2663
+ "selected_utility_gain_over_base": -0.006668480734030406,
2664
+ "selector_regret": 1.2629472265640895,
2665
+ "success_selector_gap": 0.6666666666666666,
2666
+ "success_support_gap": 0.0,
2667
+ "support_gap": 0.03412460039059321,
2668
+ "tau": 0.288910690801741
2669
+ },
2670
+ "750448296/train2": {
2671
+ "base_success": 0.0,
2672
+ "base_utility": 0.11759314934412639,
2673
+ "coverage": 0.3333333333333333,
2674
+ "execute_generated": 0.3333333333333333,
2675
+ "fallback_rate": 0.6666666666666666,
2676
+ "hidden_chart_oracle_success": 0.6666666666666666,
2677
+ "hidden_chart_oracle_utility": 1.4079964955647786,
2678
+ "outcome_ptr": 0.6666666666666666,
2679
+ "predicted_margin": 0.2956117703305941,
2680
+ "proposal_oracle_success": 0.6666666666666666,
2681
+ "proposal_oracle_utility": 1.3520947434008121,
2682
+ "selected_candidate_index": 3.0,
2683
+ "selected_success": 0.0,
2684
+ "selected_success_gain_over_base": 0.0,
2685
+ "selected_utility": 0.10535539562503497,
2686
+ "selected_utility_gain_over_base": -0.012237753719091415,
2687
+ "selector_regret": 1.246739347775777,
2688
+ "success_selector_gap": 0.6666666666666666,
2689
+ "success_support_gap": 0.0,
2690
+ "support_gap": 0.0559017521639665,
2691
+ "tau": 0.288910690801741
2692
+ },
2693
+ "767747424/train0": {
2694
+ "base_success": 0.25,
2695
+ "base_utility": 0.5731993420049548,
2696
+ "coverage": 0.5,
2697
+ "execute_generated": 0.5,
2698
+ "fallback_rate": 0.5,
2699
+ "hidden_chart_oracle_success": 0.25,
2700
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2701
+ "outcome_ptr": 0.25,
2702
+ "predicted_margin": 0.29605902981825494,
2703
+ "proposal_oracle_success": 0.25,
2704
+ "proposal_oracle_utility": 0.5464163837023079,
2705
+ "selected_candidate_index": 4.25,
2706
+ "selected_success": 0.0,
2707
+ "selected_success_gain_over_base": -0.25,
2708
+ "selected_utility": 0.08527037035673857,
2709
+ "selected_utility_gain_over_base": -0.48792897164821625,
2710
+ "selector_regret": 0.5009549129754305,
2711
+ "success_selector_gap": 0.25,
2712
+ "success_support_gap": 0.25,
2713
+ "support_gap": 0.5616356539539993,
2714
+ "tau": 0.288910690801741
2715
+ },
2716
+ "767747424/train1": {
2717
+ "base_success": 0.25,
2718
+ "base_utility": 0.5731993420049548,
2719
+ "coverage": 0.25,
2720
+ "execute_generated": 0.25,
2721
+ "fallback_rate": 0.75,
2722
+ "hidden_chart_oracle_success": 0.25,
2723
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2724
+ "outcome_ptr": 0.25,
2725
+ "predicted_margin": 0.29424454211004486,
2726
+ "proposal_oracle_success": 0.25,
2727
+ "proposal_oracle_utility": 0.5927901230752468,
2728
+ "selected_candidate_index": 2.75,
2729
+ "selected_success": 0.25,
2730
+ "selected_success_gain_over_base": 0.0,
2731
+ "selected_utility": 0.5833210749551654,
2732
+ "selected_utility_gain_over_base": 0.010121732950210571,
2733
+ "selector_regret": 0.02142970822751522,
2734
+ "success_selector_gap": 0.0,
2735
+ "success_support_gap": 0.25,
2736
+ "support_gap": 0.5152619145810604,
2737
+ "tau": 0.288910690801741
2738
+ },
2739
+ "767747424/train2": {
2740
+ "base_success": 0.25,
2741
+ "base_utility": 0.5731993420049548,
2742
+ "coverage": 0.75,
2743
+ "execute_generated": 0.75,
2744
+ "fallback_rate": 0.25,
2745
+ "hidden_chart_oracle_success": 0.25,
2746
+ "hidden_chart_oracle_utility": 0.6737305596470833,
2747
+ "outcome_ptr": 0.5,
2748
+ "predicted_margin": 0.2935865500650252,
2749
+ "proposal_oracle_success": 0.5,
2750
+ "proposal_oracle_utility": 1.0670426050201058,
2751
+ "selected_candidate_index": 3.75,
2752
+ "selected_success": 0.0,
2753
+ "selected_success_gain_over_base": -0.25,
2754
+ "selected_utility": 0.08117664931342006,
2755
+ "selected_utility_gain_over_base": -0.49202269269153476,
2756
+ "selector_regret": 0.9858659557066858,
2757
+ "success_selector_gap": 0.5,
2758
+ "success_support_gap": 0.0,
2759
+ "support_gap": 0.04100943263620138,
2760
+ "tau": 0.288910690801741
2761
+ }
2762
+ }
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/metrics_by_task.json ADDED
@@ -0,0 +1,117 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "LiftPegUpright-v1": {
3
+ "base_success": 0.2,
4
+ "base_utility": 0.9498705387115478,
5
+ "coverage": 0.8,
6
+ "execute_generated": 0.8,
7
+ "fallback_rate": 0.2,
8
+ "hidden_chart_oracle_success": 1.0,
9
+ "hidden_chart_oracle_utility": 2.0,
10
+ "outcome_ptr": 0.6,
11
+ "predicted_margin": 0.3165794975360194,
12
+ "proposal_oracle_success": 0.7333333333333333,
13
+ "proposal_oracle_utility": 1.5761187195777893,
14
+ "selected_candidate_index": 3.0,
15
+ "selected_success": 0.26666666666666666,
16
+ "selected_success_gain_over_base": 0.06666666666666667,
17
+ "selected_utility": 0.8863775124152501,
18
+ "selected_utility_gain_over_base": -0.0634930262962977,
19
+ "selector_regret": 0.7155201315879822,
20
+ "success_selector_gap": 0.4666666666666667,
21
+ "success_support_gap": 0.26666666666666666,
22
+ "support_gap": 0.4238812804222107,
23
+ "tau": 0.288910690801741
24
+ },
25
+ "PickCube-v1": {
26
+ "base_success": 0.0,
27
+ "base_utility": 0.5839591026306152,
28
+ "coverage": 1.0,
29
+ "execute_generated": 1.0,
30
+ "fallback_rate": 0.0,
31
+ "hidden_chart_oracle_success": 1.0,
32
+ "hidden_chart_oracle_utility": 2.0,
33
+ "outcome_ptr": 0.0,
34
+ "predicted_margin": 0.400859267912522,
35
+ "proposal_oracle_success": 0.0,
36
+ "proposal_oracle_utility": 0.14042885601520538,
37
+ "selected_candidate_index": 4.0,
38
+ "selected_success": 0.0,
39
+ "selected_success_gain_over_base": 0.0,
40
+ "selected_utility": 0.11226301764448483,
41
+ "selected_utility_gain_over_base": -0.4716960849861304,
42
+ "selector_regret": 0.028165838370720547,
43
+ "success_selector_gap": 0.0,
44
+ "success_support_gap": 1.0,
45
+ "support_gap": 1.8595711439847946,
46
+ "tau": 0.288910690801741
47
+ },
48
+ "PullCube-v1": {
49
+ "base_success": 0.36,
50
+ "base_utility": 0.7751704612234608,
51
+ "coverage": 0.6133333333333333,
52
+ "execute_generated": 0.6133333333333333,
53
+ "fallback_rate": 0.38666666666666666,
54
+ "hidden_chart_oracle_success": 0.52,
55
+ "hidden_chart_oracle_utility": 1.1467878860235214,
56
+ "outcome_ptr": 0.4533333333333333,
57
+ "predicted_margin": 0.3093096789241044,
58
+ "proposal_oracle_success": 0.48,
59
+ "proposal_oracle_utility": 1.0064672262904544,
60
+ "selected_candidate_index": 3.0,
61
+ "selected_success": 0.30666666666666664,
62
+ "selected_success_gain_over_base": -0.05333333333333334,
63
+ "selected_utility": 0.667844362364461,
64
+ "selected_utility_gain_over_base": -0.10732609885899971,
65
+ "selector_regret": 0.46678621487071115,
66
+ "success_selector_gap": 0.24,
67
+ "success_support_gap": 0.09333333333333334,
68
+ "support_gap": 0.23335681801661848,
69
+ "tau": 0.28891069080174125
70
+ },
71
+ "PushCube-v1": {
72
+ "base_success": 0.16666666666666666,
73
+ "base_utility": 0.3924041787783305,
74
+ "coverage": 0.8333333333333334,
75
+ "execute_generated": 0.8333333333333334,
76
+ "fallback_rate": 0.16666666666666666,
77
+ "hidden_chart_oracle_success": 0.8333333333333334,
78
+ "hidden_chart_oracle_utility": 1.7015800600250561,
79
+ "outcome_ptr": 0.7777777777777778,
80
+ "predicted_margin": 0.3378348162819955,
81
+ "proposal_oracle_success": 0.6666666666666666,
82
+ "proposal_oracle_utility": 1.364444797237714,
83
+ "selected_candidate_index": 1.9444444444444444,
84
+ "selected_success": 0.3888888888888889,
85
+ "selected_success_gain_over_base": 0.2222222222222222,
86
+ "selected_utility": 0.8178884767823749,
87
+ "selected_utility_gain_over_base": 0.4254842980040444,
88
+ "selector_regret": 0.5481036975979805,
89
+ "success_selector_gap": 0.2777777777777778,
90
+ "success_support_gap": 0.3333333333333333,
91
+ "support_gap": 0.6355552027622858,
92
+ "tau": 0.2889106908017409
93
+ },
94
+ "StackCube-v1": {
95
+ "base_success": 0.2727272727272727,
96
+ "base_utility": 0.7763947045261209,
97
+ "coverage": 0.8787878787878788,
98
+ "execute_generated": 0.8787878787878788,
99
+ "fallback_rate": 0.12121212121212122,
100
+ "hidden_chart_oracle_success": 1.0,
101
+ "hidden_chart_oracle_utility": 2.0,
102
+ "outcome_ptr": 0.6060606060606061,
103
+ "predicted_margin": 0.33195146262815795,
104
+ "proposal_oracle_success": 0.45454545454545453,
105
+ "proposal_oracle_utility": 1.1071163348176263,
106
+ "selected_candidate_index": 3.242424242424242,
107
+ "selected_success": 0.12121212121212122,
108
+ "selected_success_gain_over_base": -0.15151515151515152,
109
+ "selected_utility": 0.43286885501760425,
110
+ "selected_utility_gain_over_base": -0.3435258495085167,
111
+ "selector_regret": 0.6748905412175439,
112
+ "success_selector_gap": 0.3333333333333333,
113
+ "success_support_gap": 0.5454545454545454,
114
+ "support_gap": 0.8928836651823737,
115
+ "tau": 0.28891069080174087
116
+ }
117
+ }
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/report.md ADDED
@@ -0,0 +1,17 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Learned Dominance-Calibrated CTT Selector
2
+
3
+ Calibration rows: `432`
4
+ Eval rows: `144`
5
+ Selected ridge lambda: `100.0`
6
+ Tau: `0.288911`
7
+ Feature set: `basic`
8
+ Target: `success`
9
+
10
+ The ridge calibrator and threshold are fit on calibration measured rows only. Eval outcomes are used only for reporting.
11
+
12
+ | Split | Coverage | Fallback | Base success | Selected success | Proposal oracle | OutcomePTR | Success support gap | Success selector gap |
13
+ | --- | ---: | ---: | ---: | ---: | ---: | ---: | ---: | ---: |
14
+ | calibration | 0.7176 | 0.2824 | 0.2778 | 0.3125 | 0.4722 | 0.4907 | 0.2315 | 0.2014 |
15
+ | eval | 0.7292 | 0.2708 | 0.2917 | 0.2639 | 0.5139 | 0.5347 | 0.2639 | 0.2847 |
16
+
17
+ This is a selector diagnostic over already measured candidates, not a new rollout.
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/split_hash.txt ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "calibration_target_index": {
3
+ "content_hash": "1d15143588697e89f7fc2f6375b2745e4d479c726c9a947ff7eb7b2705280e1a",
4
+ "retrieval_index_allowed": true,
5
+ "split": "train",
6
+ "split_hash": "867a94a1c67686c082765671a7bf8d74a9d633085e1b2619c83197c60bf58372"
7
+ },
8
+ "eval_target_index": {
9
+ "content_hash": "d258ed27d587e6f41ef6ffd520832087fe7626b5e5cba2bd6352f50a1f45125a",
10
+ "retrieval_index_allowed": false,
11
+ "split": "test",
12
+ "split_hash": "84151df9e88c219b76751677558cc5bbbc19082bf358cea7330158e9e403e25f"
13
+ }
14
+ }
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/table.tex ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ % Auto-generated by scripts/eval_learned_dominance_selector.py
2
+ \begin{tabular}{lrrrrrrrr}
3
+ \toprule
4
+ Rows & Coverage & Fallback & Base succ. & Selected succ. & Oracle succ. & OutcomePTR & Succ. support gap & Succ. selector gap \\
5
+ \midrule
6
+ 144 & 0.7292 & 0.2708 & 0.2917 & 0.2639 & 0.5139 & 0.5347 & 0.2639 & 0.2847 \\
7
+ \bottomrule
8
+ \end{tabular}
runs/ctt_base_context_obs_learned_dominance_success_train_to_test/train.log ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ trained ridge dominance calibrator on calibration measured rows only
2
+ selected_lambda=100.0
3
+ tau=0.288911
4
+ calibration_input=runs/ctt_base_context_obs_train_cal_rollout_comparison/combined_measured_candidates.json