1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
|
{
"author_history": "/scratch/yurenh2/sdil-contrastive-bias/dualprop/runs/dp-bias-b1-activity-r1-raw/2026_08_06_12_37_12/hist.npy",
"cell_id": "activity-r1-raw",
"command": [
"/scratch/yurenh2/venvs/dualprop/bin/python",
"train.py",
"--model",
"miniCNN",
"--dataset",
"cifar10",
"--num-epochs",
"20",
"--batch-size",
"100",
"--learning-rate",
"0.025",
"--learning-rate-final",
"0.025",
"--warmup-learning-rate",
"0.025",
"--warmup-epochs",
"0",
"--decay-epochs",
"20",
"--momentum",
"0.9",
"--weight-decay",
"5e-4",
"--dtype",
"float32",
"--param-dtype",
"float32",
"--percent-train",
"90",
"--percent-val",
"10",
"--seeds",
"1988",
"--feedback-seed",
"1729",
"--gradient-diagnostics",
"none",
"--spectral-diagnostics",
"none",
"--test-policy",
"none",
"--early-stop-policy",
"none",
"--learning-algorithm",
"dualprop-lagr-ff",
"--experiment-name",
"dp-bias-b1-activity-r1-raw",
"--optimizer-schedule",
"author",
"--loss",
"sce",
"--alpha",
"0.0",
"--beta",
"0.1",
"--inference-sequence",
"fwK",
"--inference-passes-nudged",
"16",
"--dp-bias-kind",
"activity",
"--dp-bias-rule",
"raw",
"--dp-bias-ratio",
"1.0",
"--dp-bias-seed",
"6100",
"--dp-bias-calibration-examples",
"64"
],
"completed_unix_time": 1786037864.6565628,
"driver_wall_seconds": 41.14709734916687,
"experiment_name": "dp-bias-b1-activity-r1-raw",
"hardware": {
"cuda_visible_devices": "7",
"memory_total_mib": 8192,
"name": "NVIDIA GeForce GTX 1080",
"physical_index": 7,
"uuid": "GPU-df3d5933-c3e9-9ea1-d595-9d2a3b68c14b"
},
"history": {
"best_epoch": 1,
"best_validation_accuracy": 9.399999618530273,
"curves": {
"instruction_observations_for_predictor": [
0.0
],
"maximum_used_clean_difference_relative_error": [
NaN
],
"neutral_observations": [
100.0
],
"post_bias_raw_bias_rms_ratio": [
NaN
],
"raw_bias_clean_difference_rms_ratio": [
NaN
],
"train_accuracy": [
10.08666706085205
],
"train_loss": [
NaN
],
"train_time": [
18.427337884902954
],
"used_clean_difference_rms_ratio": [
NaN
],
"val_accuracy": [
9.399999618530273
],
"val_loss": [
NaN
],
"val_time": [
0.8281819820404053
]
},
"dp_bias_initialization": {
"bias_kind": "activity",
"bias_ratio_target": 1.0,
"calibration_examples": 64,
"instruction_observations_for_predictor": 0,
"realized_bias_difference_rms_ratio": [
1.0000001192092896,
1.0,
1.0
]
},
"epochs_completed": 1,
"final_validation_accuracy": 9.399999618530273,
"finite": false,
"test_accuracy": NaN
},
"kind": "activity",
"output": "/home/yurenh2/sdil/results/contrastive_bias/b1/dp-bias-b1-activity-r1-raw.json",
"ratio": 1.0,
"registry_sha256": "6ce612c3890f395be7f77684204a19933aa83c0c92e51df991ef36258ae054b3",
"return_code": 0,
"rule": "raw",
"source": {
"author_commit": "79169abac635715e2167f1bb8a089f239d06c431",
"author_upstream": "7b2595b34421e1483a721dbfdeff8cdabda3a1ff",
"sdil_commit": "05591963386af3d5f8ccd348ada45c393a3996d4",
"tracked_files": {
"CONTRASTIVE_BIAS.md": "b6dad5734efc8f04945f49ffce9c9ddae8c82fa28bff6cbdb37ef08287b28ffc",
"experiments/analyze_contrastive_bias_b1.py": "945f5741b2676dc2d75e900c473a10c641eb80d701881533116b60b5582fedee",
"experiments/contrastive_bias_b1.py": "964408b7dc064e69bf348a9a389aaea0e749735a31ca5c6db96d2a2660db5325",
"external/dualprop_patches/0020-experiment-add-neuron-specific-bias-to-Dual-Prop.patch": "7ae57542682482dc17f7073ed32c16876e60826431b2d4fbaa391d68332c38f0"
}
},
"stage": "contrastive_bias_b1",
"status": "completed",
"timeout_seconds": 7200
}
|
