Upload benchmarks/intervention_results.json with huggingface_hub
Browse files
benchmarks/intervention_results.json
ADDED
|
@@ -0,0 +1,148 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"structures": {
|
| 3 |
+
"confounded": {
|
| 4 |
+
"rmse": 0.681593135140375,
|
| 5 |
+
"direction_accuracy": 0.0,
|
| 6 |
+
"trajectory_correlation": 0.0,
|
| 7 |
+
"null_detected": 0.0,
|
| 8 |
+
"pred_cate_mean": 0.6810493767981727,
|
| 9 |
+
"true_cate_mean": 0.0,
|
| 10 |
+
"n_seeds": 5
|
| 11 |
+
},
|
| 12 |
+
"mediated": {
|
| 13 |
+
"rmse": 0.8923469699693654,
|
| 14 |
+
"direction_accuracy": 0.4666666666666667,
|
| 15 |
+
"trajectory_correlation": -0.1076186800934075,
|
| 16 |
+
"null_detected": "N/A",
|
| 17 |
+
"pred_cate_mean": -0.0896030426301683,
|
| 18 |
+
"true_cate_mean": 0.5514786841923025,
|
| 19 |
+
"n_seeds": 5
|
| 20 |
+
},
|
| 21 |
+
"time_varying_confounded": {
|
| 22 |
+
"rmse": 0.40978652111886243,
|
| 23 |
+
"direction_accuracy": 1.0,
|
| 24 |
+
"trajectory_correlation": 0.0,
|
| 25 |
+
"null_detected": "N/A",
|
| 26 |
+
"pred_cate_mean": 0.18326412199096132,
|
| 27 |
+
"true_cate_mean": 0.5911041763020383,
|
| 28 |
+
"n_seeds": 5
|
| 29 |
+
},
|
| 30 |
+
"feedback": {
|
| 31 |
+
"rmse": 0.31327788634616005,
|
| 32 |
+
"direction_accuracy": 1.0,
|
| 33 |
+
"trajectory_correlation": 0.0,
|
| 34 |
+
"null_detected": "N/A",
|
| 35 |
+
"pred_cate_mean": 0.5442088283598423,
|
| 36 |
+
"true_cate_mean": 0.514855316018063,
|
| 37 |
+
"n_seeds": 5
|
| 38 |
+
},
|
| 39 |
+
"instrumental_variable": {
|
| 40 |
+
"rmse": 0.70705308193879,
|
| 41 |
+
"direction_accuracy": 0.9566666666666667,
|
| 42 |
+
"trajectory_correlation": -0.021855126454785802,
|
| 43 |
+
"null_detected": "N/A",
|
| 44 |
+
"pred_cate_mean": 0.2339213555591802,
|
| 45 |
+
"true_cate_mean": 0.877110000875206,
|
| 46 |
+
"n_seeds": 5
|
| 47 |
+
},
|
| 48 |
+
"non_identifiable": {
|
| 49 |
+
"rmse": 0.29658983764206714,
|
| 50 |
+
"direction_accuracy": 1.0,
|
| 51 |
+
"trajectory_correlation": -0.27477913370448287,
|
| 52 |
+
"null_detected": "N/A",
|
| 53 |
+
"pred_cate_mean": 0.8563931959867478,
|
| 54 |
+
"true_cate_mean": 0.6002255928070188,
|
| 55 |
+
"n_seeds": 5
|
| 56 |
+
}
|
| 57 |
+
},
|
| 58 |
+
"scenarios": {
|
| 59 |
+
"step": {
|
| 60 |
+
"scenario": "step",
|
| 61 |
+
"description": "Step Intervention",
|
| 62 |
+
"n_seeds": 5,
|
| 63 |
+
"trajectory_rmse": {
|
| 64 |
+
"mean": 0.44790337647382295,
|
| 65 |
+
"std": 0.11992705799993689
|
| 66 |
+
},
|
| 67 |
+
"ate_error": {
|
| 68 |
+
"mean": 0.1329101725025971,
|
| 69 |
+
"std": 0.13308464275986506
|
| 70 |
+
},
|
| 71 |
+
"direction_accuracy": {
|
| 72 |
+
"mean": 0.5377777777777777,
|
| 73 |
+
"std": 0.025915341754868006
|
| 74 |
+
},
|
| 75 |
+
"time_s": 1.6741529230494052,
|
| 76 |
+
"extra": {}
|
| 77 |
+
},
|
| 78 |
+
"dose_response": {
|
| 79 |
+
"scenario": "dose_response",
|
| 80 |
+
"description": "Dose-Response Curve",
|
| 81 |
+
"n_seeds": 5,
|
| 82 |
+
"trajectory_rmse": {
|
| 83 |
+
"mean": 0.2758524529939483,
|
| 84 |
+
"std": 0.07663341940224565
|
| 85 |
+
},
|
| 86 |
+
"ate_error": {
|
| 87 |
+
"mean": 0.13401241106339618,
|
| 88 |
+
"std": 0.14771866074543336
|
| 89 |
+
},
|
| 90 |
+
"direction_accuracy": {
|
| 91 |
+
"mean": 0.5444444444444444,
|
| 92 |
+
"std": 0.03296088216486962
|
| 93 |
+
},
|
| 94 |
+
"time_s": 324.4969568946399,
|
| 95 |
+
"extra": {}
|
| 96 |
+
},
|
| 97 |
+
"policy": {
|
| 98 |
+
"scenario": "policy",
|
| 99 |
+
"description": "Policy Comparison",
|
| 100 |
+
"n_seeds": 5,
|
| 101 |
+
"trajectory_rmse": {
|
| 102 |
+
"mean": 0.09002535829029422,
|
| 103 |
+
"std": 0.09175816283732381
|
| 104 |
+
},
|
| 105 |
+
"ate_error": {
|
| 106 |
+
"mean": 0.0615661122289803,
|
| 107 |
+
"std": 0.06800788225181435
|
| 108 |
+
},
|
| 109 |
+
"direction_accuracy": {
|
| 110 |
+
"mean": 0.6,
|
| 111 |
+
"std": 0.48989794855663565
|
| 112 |
+
},
|
| 113 |
+
"time_s": 181.22026548418216,
|
| 114 |
+
"extra": {}
|
| 115 |
+
},
|
| 116 |
+
"timing": {
|
| 117 |
+
"scenario": "timing",
|
| 118 |
+
"description": "Intervention Timing",
|
| 119 |
+
"n_seeds": 5,
|
| 120 |
+
"trajectory_rmse": {
|
| 121 |
+
"mean": 0.42171894065880655,
|
| 122 |
+
"std": 0.1268720400592207
|
| 123 |
+
},
|
| 124 |
+
"ate_error": {
|
| 125 |
+
"mean": 0.23317682692018069,
|
| 126 |
+
"std": 0.07347633462864078
|
| 127 |
+
},
|
| 128 |
+
"direction_accuracy": {
|
| 129 |
+
"mean": 0.5183333333333333,
|
| 130 |
+
"std": 0.04718462555630688
|
| 131 |
+
},
|
| 132 |
+
"time_s": 14.694185567088425,
|
| 133 |
+
"extra": {}
|
| 134 |
+
}
|
| 135 |
+
},
|
| 136 |
+
"machine": {
|
| 137 |
+
"timestamp": "2026-04-09T05:03:17.988416+00:00",
|
| 138 |
+
"hostname": "7aa1a49e1d2f",
|
| 139 |
+
"platform": "Linux-6.8.0-106-generic-x86_64-with-glibc2.39",
|
| 140 |
+
"python": "3.12.13",
|
| 141 |
+
"cpu": "x86_64",
|
| 142 |
+
"cpu_count": 384,
|
| 143 |
+
"torch_version": "2.10.0+cu130",
|
| 144 |
+
"gpu": "NVIDIA GeForce RTX 5090",
|
| 145 |
+
"gpu_count": 1,
|
| 146 |
+
"ram_gb": 540.5
|
| 147 |
+
}
|
| 148 |
+
}
|