Kurtosis-calibrated gradient clipping / results.json
Beats tuned baseline
1{
2 "math_check": {
3 "max_inverse_error": 1.3877787807814457e-17,
4 "minimum_projected_tail_margin": 0.0,
5 "rows": [
6 {
7 "k": 1.01,
8 "c": 1.0512492197250387,
9 "transition_polynomial": -4.440892098500626e-16
10 },
11 {
12 "k": 1.1,
13 "c": 1.1705367195742484,
14 "transition_polynomial": 4.440892098500626e-16
15 },
16 {
17 "k": 2,
18 "c": 1.618033988749895,
19 "transition_polynomial": 8.881784197001252e-16
20 },
21 {
22 "k": 3,
23 "c": 1.9318516525781366,
24 "transition_polynomial": 0.0
25 },
26 {
27 "k": 10,
28 "c": 3.302775637731995,
29 "transition_polynomial": 1.4210854715202004e-14
30 },
31 {
32 "k": 100,
33 "c": 10.04938296807921,
34 "transition_polynomial": 0.0
35 }
36 ]
37 },
38 "empirical_tail_check": [
39 {
40 "distribution": "student_t",
41 "nu": 5,
42 "kurtosis_bound": 9.450000000000001,
43 "delta": 0.01,
44 "threshold": 5.470207866221638,
45 "observed_tail": 0.000436,
46 "bound_formula_at_threshold": 0.010000000000000002
47 },
48 {
49 "distribution": "student_t",
50 "nu": 5,
51 "kurtosis_bound": 9.450000000000001,
52 "delta": 0.05,
53 "threshold": 3.6974089100681917,
54 "observed_tail": 0.0024545,
55 "bound_formula_at_threshold": 0.05000000000000001
56 },
57 {
58 "distribution": "student_t",
59 "nu": 8,
60 "kurtosis_bound": 4.7250000000000005,
61 "delta": 0.01,
62 "threshold": 4.4948320661837675,
63 "observed_tail": 0.0003925,
64 "bound_formula_at_threshold": 0.010000000000000004
65 },
66 {
67 "distribution": "student_t",
68 "nu": 8,
69 "kurtosis_bound": 4.7250000000000005,
70 "delta": 0.05,
71 "threshold": 3.068026708671015,
72 "observed_tail": 0.003743,
73 "bound_formula_at_threshold": 0.05000000000000004
74 },
75 {
76 "distribution": "student_t",
77 "nu": 20,
78 "kurtosis_bound": 3.54375,
79 "delta": 0.01,
80 "threshold": 4.107212175049988,
81 "observed_tail": 0.0001855,
82 "bound_formula_at_threshold": 0.010000000000000005
83 },
84 {
85 "distribution": "student_t",
86 "nu": 20,
87 "kurtosis_bound": 3.54375,
88 "delta": 0.05,
89 "threshold": 2.819941140811586,
90 "observed_tail": 0.0037685,
91 "bound_formula_at_threshold": 0.05
92 }
93 ],
94 "optimizer": {
95 "none": {
96 "final_loss": 0.0013225229805502608,
97 "mean_last50_loss": 0.0013702425855572865,
98 "max_loss": 27.038855599758563,
99 "spike_steps": 238.2,
100 "mean_tail_above_threshold": null,
101 "final_abs_x": 0.045344010240534596,
102 "replicate_std_final_loss": 0.0010346012527671547
103 },
104 "fixed": {
105 "final_loss": 5.07767451269562e-05,
106 "mean_last50_loss": 9.46825981257205e-05,
107 "max_loss": 30.764175,
108 "spike_steps": 248.6,
109 "mean_tail_above_threshold": 0.15234375,
110 "final_abs_x": 0.00791960988221092,
111 "replicate_std_final_loss": 8.242405461179995e-05
112 },
113 "calibrated": {
114 "final_loss": 0.0013225229805483966,
115 "mean_last50_loss": 0.0013702425855083417,
116 "max_loss": 27.05331330396915,
117 "spike_steps": 6.9,
118 "mean_tail_above_threshold": 0.00065625,
119 "final_abs_x": 0.045344010240479314,
120 "replicate_std_final_loss": 0.0010346012527644123
121 }
122 }
123}