logbook-files/baseline_ablation_results.csv

step,baseline_accuracy,nobaseline_accuracy,baseline_asr,nobaseline_asr,baseline_attacker_reward,nobaseline_attacker_reward,baseline_defender_reward,nobaseline_defender_reward,baseline_response_length,nobaseline_response_length
0,0.3535,0.357,0.5035,0.507,0.2035,0.207,-0.3965,-0.393,532.9,526.6
1,0.3497,0.3388,0.4847,0.487,0.1798,0.1885,-0.3953,-0.4103,544.2,533.2
2,0.3728,0.3749,0.4932,0.5213,0.1834,0.2243,-0.3671,-0.3734,535.1,529.8
3,0.3576,0.3344,0.4636,0.4791,0.149,0.1836,-0.3774,-0.4131,537.8,531.6
4,0.3637,0.3368,0.4557,0.4797,0.1362,0.1858,-0.3664,-0.4099,538.2,534.3
5,0.3827,0.3652,0.4609,0.5064,0.1367,0.214,-0.3425,-0.3806,539.1,533.8
6,0.3953,0.3809,0.4599,0.5204,0.1311,0.2295,-0.3252,-0.364,536.6,537.6
7,0.3983,0.3778,0.4497,0.5155,0.1163,0.2262,-0.3175,-0.3663,542.3,528.3
8,0.4105,0.3932,0.4489,0.5291,0.1109,0.2413,-0.3006,-0.3501,542.9,527.9
9,0.3946,0.3526,0.4203,0.4866,0.0778,0.2003,-0.312,-0.3898,551.2,536.0
10,0.4179,0.3905,0.4311,0.5225,0.0841,0.2378,-0.2842,-0.3511,550.0,535.5
11,0.4274,0.4011,0.4284,0.5311,0.077,0.2479,-0.2702,-0.3397,545.4,536.2
12,0.4291,0.3962,0.4181,0.5241,0.0623,0.2424,-0.2641,-0.3437,549.6,541.3
13,0.4279,0.3859,0.4052,0.5115,0.0452,0.2313,-0.2609,-0.3532,551.5,529.5
14,0.4243,0.3707,0.39,0.4939,0.0257,0.2152,-0.2603,-0.3676,545.9,530.5
15,0.4293,0.3731,0.3838,0.4939,0.0153,0.2167,-0.2509,-0.3643,543.7,533.9
16,0.4443,0.3954,0.3877,0.5135,0.015,0.2379,-0.2318,-0.3411,548.2,535.7
17,0.4384,0.3763,0.371,0.4916,-0.0058,0.2174,-0.2335,-0.3595,547.3,544.7
18,0.454,0.4003,0.376,0.5127,-0.0048,0.24,-0.2138,-0.3346,553.0,533.6
19,0.4606,0.4065,0.3722,0.5158,-0.0127,0.2444,-0.2032,-0.3276,546.9,537.1
20,0.4716,0.4216,0.373,0.5276,-0.0158,0.2577,-0.1882,-0.3117,554.5,540.2
21,0.4684,0.4084,0.3598,0.5109,-0.033,0.2424,-0.1875,-0.3241,558.1,543.9
22,0.4613,0.3876,0.3429,0.4866,-0.0537,0.2194,-0.1907,-0.344,546.4,537.3
23,0.4665,0.3915,0.3385,0.4867,-0.0619,0.2208,-0.1817,-0.3393,549.6,547.1
24,0.4859,0.4239,0.3485,0.5151,-0.0558,0.2505,-0.1586,-0.3061,551.6,543.1
25,0.4779,0.4018,0.3313,0.4889,-0.0767,0.2254,-0.1628,-0.3273,561.1,540.5
26,0.4786,0.3972,0.323,0.4799,-0.0887,0.2176,-0.1584,-0.3312,551.5,542.4
27,0.4825,0.3989,0.3179,0.4772,-0.0974,0.216,-0.1509,-0.3286,557.0,547.8
28,0.4897,0.4076,0.3165,0.4812,-0.1024,0.2211,-0.1401,-0.3191,551.9,549.6
29,0.4962,0.415,0.3145,0.4837,-0.108,0.2245,-0.1301,-0.3109,550.4,535.6
30,0.4899,0.3968,0.2998,0.4605,-0.1262,0.2022,-0.1329,-0.3283,558.8,547.1
31,0.5013,0.4142,0.3031,0.4727,-0.1264,0.2152,-0.1181,-0.3101,550.7,541.2
32,0.5191,0.4446,0.3129,0.4979,-0.12,0.2411,-0.0968,-0.2788,558.0,550.3
33,0.5192,0.4396,0.3051,0.4874,-0.1311,0.2312,-0.0934,-0.283,550.6,546.7
34,0.5181,0.4324,0.2964,0.4745,-0.1432,0.2189,-0.0911,-0.2895,558.7,540.1
35,0.5204,0.432,0.2911,0.4684,-0.1518,0.2132,-0.0856,-0.289,552.6,542.9
36,0.519,0.4243,0.2823,0.4549,-0.1639,0.2,-0.0838,-0.2959,565.5,549.7
37,0.5174,0.4165,0.2735,0.441,-0.1759,0.1864,-0.0822,-0.303,558.9,539.4
38,0.5368,0.4505,0.2857,0.469,-0.1668,0.2144,-0.0597,-0.2681,564.7,541.3
39,0.533,0.4385,0.275,0.4508,-0.1807,0.1962,-0.0604,-0.2794,561.0,539.3
40,0.5391,0.4462,0.2743,0.4523,-0.1845,0.1975,-0.0512,-0.2709,560.1,548.8
41,0.5362,0.4362,0.2647,0.436,-0.1971,0.1808,-0.0511,-0.2801,552.1,542.1
42,0.5263,0.4122,0.2483,0.4056,-0.2166,0.15,-0.058,-0.3033,566.2,550.5
43,0.5495,0.4543,0.265,0.4414,-0.2029,0.1852,-0.0319,-0.2604,556.9,538.2
44,0.5534,0.4581,0.2626,0.4388,-0.2082,0.1818,-0.0251,-0.2558,566.6,538.7
45,0.5463,0.44,0.2494,0.4143,-0.2244,0.1564,-0.0293,-0.2732,553.5,548.9
46,0.5559,0.4554,0.253,0.4233,-0.2236,0.1643,-0.0169,-0.257,554.4,544.6
47,0.5531,0.4461,0.2443,0.4076,-0.2352,0.1473,-0.0169,-0.2656,556.5,550.6
48,0.5525,0.4412,0.2379,0.3964,-0.2444,0.1347,-0.0148,-0.2697,555.7,545.6
49,0.5627,0.4579,0.2423,0.4068,-0.2427,0.1436,-0.0019,-0.2522,555.5,542.1
50,0.5718,0.4725,0.2459,0.4153,-0.2419,0.1502,0.0099,-0.2368,562.1,544.0
51,0.5623,0.45,0.2309,0.3866,-0.2596,0.1197,0.003,-0.2586,554.1,540.7
52,0.5601,0.4424,0.2234,0.3729,-0.2697,0.1039,0.0034,-0.2655,561.0,540.7
53,0.5595,0.4378,0.2175,0.3625,-0.2782,0.0911,0.0053,-0.2693,553.1,546.6
54,0.562,0.4396,0.2148,0.3584,-0.2835,0.0846,0.0103,-0.2668,565.5,550.0
55,0.5602,0.433,0.208,0.3461,-0.2928,0.0697,0.0111,-0.2726,555.3,547.0
56,0.566,0.4414,0.2089,0.3491,-0.2945,0.0699,0.0193,-0.2635,553.5,550.8
57,0.577,0.4605,0.2151,0.3628,-0.2908,0.0807,0.0328,-0.2436,558.4,548.4
58,0.585,0.4736,0.2183,0.3707,-0.29,0.0854,0.0432,-0.2298,554.0,537.9
59,0.5777,0.456,0.2063,0.3481,-0.3044,0.0596,0.0382,-0.2467,557.2,543.3
60,0.5871,0.4721,0.2112,0.3594,-0.3019,0.0675,0.0499,-0.2299,560.7,550.8
61,0.5733,0.4417,0.1929,0.3245,-0.3226,0.0289,0.0384,-0.2595,556.3,540.9
62,0.5943,0.4811,0.2096,0.3594,-0.3083,0.0602,0.0617,-0.2194,553.7,538.4
63,0.5859,0.4616,0.1968,0.3358,-0.3233,0.0327,0.0555,-0.2382,555.9,539.5
64,0.5828,0.4528,0.1895,0.323,-0.3329,0.016,0.0546,-0.2463,563.1,541.7
65,0.5999,0.4845,0.2025,0.351,-0.3222,0.0399,0.0738,-0.2139,557.1,535.6
66,0.6051,0.4925,0.2037,0.3556,-0.3232,0.0402,0.0812,-0.2051,561.0,549.0
67,0.605,0.4899,0.1996,0.3498,-0.3294,0.0301,0.0832,-0.207,560.8,536.7
68,0.5957,0.469,0.1865,0.3259,-0.3447,0.0018,0.076,-0.2272,550.9,539.9
69,0.5867,0.4487,0.1737,0.3029,-0.3597,-0.0257,0.069,-0.2468,552.9,548.3
70,0.5935,0.4601,0.1767,0.3118,-0.3587,-0.0213,0.0778,-0.2347,558.5,540.0
71,0.5991,0.469,0.1786,0.3185,-0.3589,-0.0192,0.0854,-0.2251,560.6,547.8
72,0.604,0.4767,0.1799,0.3242,-0.3596,-0.0181,0.0922,-0.2168,556.5,534.8
73,0.599,0.4648,0.1715,0.3106,-0.37,-0.0365,0.0893,-0.228,559.8,540.8
74,0.6067,0.478,0.1757,0.3224,-0.3678,-0.0294,0.0988,-0.2141,556.0,532.5
75,0.6092,0.4811,0.1749,0.3243,-0.3706,-0.0322,0.1032,-0.2103,551.7,544.1
76,0.6,0.4608,0.1624,0.3029,-0.3851,-0.0582,0.0959,-0.23,558.1,532.1
77,0.6021,0.4631,0.1613,0.3045,-0.3881,-0.0613,0.0998,-0.2269,552.0,540.6
78,0.6048,0.4666,0.1607,0.3074,-0.3905,-0.063,0.1043,-0.2228,544.3,543.4
79,0.6064,0.468,0.1593,0.3086,-0.3938,-0.0664,0.1077,-0.2207,550.8,538.9
80,0.6121,0.4776,0.1619,0.3181,-0.393,-0.0614,0.1151,-0.2105,550.1,537.0
81,0.6263,0.5043,0.1731,0.3449,-0.3836,-0.039,0.131,-0.1831,543.8,539.4
82,0.6102,0.4704,0.154,0.3113,-0.4044,-0.0769,0.1166,-0.2164,546.0,538.3
83,0.6129,0.4744,0.154,0.3157,-0.4063,-0.0767,0.1211,-0.2118,544.2,539.0
84,0.6085,0.4639,0.1467,0.3059,-0.4153,-0.0905,0.1183,-0.2216,545.7,542.2
85,0.6328,0.5109,0.1683,0.3537,-0.3954,-0.0466,0.1442,-0.1739,539.3,530.0
86,0.6157,0.4752,0.1485,0.319,-0.4169,-0.0851,0.1288,-0.209,551.6,539.5
87,0.6322,0.5068,0.1623,0.3516,-0.4047,-0.056,0.1469,-0.1768,542.4,528.2
88,0.6322,0.5053,0.1598,0.3514,-0.4089,-0.0596,0.1484,-0.1776,546.8,534.5
89,0.6372,0.5138,0.1622,0.3612,-0.4081,-0.053,0.1549,-0.1685,545.4,524.9
90,0.634,0.5058,0.1565,0.3549,-0.4154,-0.0622,0.1533,-0.176,539.4,534.3
91,0.6382,0.5121,0.1582,0.3632,-0.4153,-0.0565,0.1589,-0.1692,536.3,525.5
92,0.6184,0.4708,0.1361,0.3239,-0.4389,-0.0983,0.1407,-0.21,541.9,527.4
93,0.6319,0.4959,0.1472,0.3512,-0.4293,-0.0732,0.1556,-0.1844,543.7,525.9
94,0.6336,0.4976,0.1466,0.355,-0.4314,-0.0713,0.1588,-0.1822,536.2,529.6
95,0.6414,0.5114,0.1522,0.371,-0.4274,-0.057,0.1679,-0.168,544.3,523.2
96,0.6303,0.4875,0.1389,0.3492,-0.4421,-0.0801,0.1583,-0.1914,535.1,521.4
97,0.64,0.5051,0.1464,0.369,-0.4361,-0.0614,0.1693,-0.1734,539.8,524.7
98,0.6402,0.5038,0.1444,0.3698,-0.4395,-0.0614,0.1708,-0.1742,537.8,526.7
99,0.6228,0.4674,0.125,0.3354,-0.4603,-0.0961,0.1548,-0.2102,530.0,533.1
100,0.646,0.5123,0.1462,0.3823,-0.4405,-0.0494,0.1794,-0.1648,536.4,531.8
101,0.6389,0.4963,0.137,0.3683,-0.451,-0.0632,0.1735,-0.1803,529.8,519.3
102,0.6329,0.4828,0.1291,0.3566,-0.4603,-0.0743,0.1688,-0.1934,540.4,527.5
103,0.6475,0.5105,0.1418,0.3861,-0.4489,-0.044,0.1847,-0.1652,539.8,516.8
104,0.6331,0.4801,0.1255,0.3573,-0.4665,-0.0716,0.1715,-0.1952,527.2,523.5
105,0.6498,0.512,0.1403,0.3907,-0.453,-0.0367,0.1894,-0.1629,531.1,519.2
106,0.6323,0.4756,0.1211,0.3557,-0.4735,-0.0699,0.1731,-0.1988,538.4,521.1
107,0.6536,0.5168,0.1406,0.3981,-0.4552,-0.0253,0.1957,-0.1572,536.0,519.5
108,0.6306,0.4693,0.1159,0.3516,-0.4812,-0.0692,0.1738,-0.2042,534.3,509.4
109,0.654,0.5147,0.1376,0.3979,-0.4608,-0.0201,0.1984,-0.1584,534.7,515.5
110,0.6517,0.5088,0.1336,0.3927,-0.4659,-0.0222,0.1972,-0.1639,526.1,512.6
111,0.6349,0.4737,0.1151,0.3582,-0.4856,-0.0532,0.1815,-0.1986,534.6,511.9
112,0.6385,0.4795,0.1171,0.3643,-0.4848,-0.0434,0.1862,-0.1924,533.2,504.1
113,0.6412,0.4837,0.1183,0.3686,-0.4847,-0.035,0.19,-0.1878,532.3,498.8
114,0.6349,0.4697,0.1104,0.3545,-0.4938,-0.0448,0.1847,-0.2014,522.8,498.6
115,0.6565,0.5118,0.1306,0.3963,-0.4747,0.0016,0.2075,-0.1589,533.9,496.1
116,0.6516,0.5006,0.1242,0.3847,-0.4822,-0.0051,0.2036,-0.1696,524.8,504.9
117,0.6496,0.4954,0.1207,0.3788,-0.4868,-0.0059,0.2026,-0.1745,532.5,490.1
118,0.6611,0.5171,0.1308,0.3996,-0.4778,0.0202,0.2151,-0.1524,520.6,501.1
119,0.6441,0.4819,0.1124,0.3633,-0.4972,-0.0106,0.1991,-0.1872,519.1,491.5
120,0.6563,0.505,0.1232,0.3851,-0.4875,0.017,0.2122,-0.1637,521.7,492.4
121,0.6514,0.4942,0.117,0.3727,-0.4947,0.0105,0.2084,-0.1742,522.4,490.3
122,0.6456,0.4813,0.1099,0.3582,-0.5028,0.002,0.2035,-0.1866,526.8,480.0
123,0.6629,0.5149,0.1259,0.3898,-0.4878,0.0398,0.2218,-0.1527,523.9,489.1
124,0.6589,0.5057,0.1206,0.3785,-0.4941,0.0348,0.2186,-0.1616,525.5,482.2
125,0.6426,0.472,0.1032,0.3425,-0.5125,0.0053,0.2033,-0.1948,525.1,480.0
126,0.6492,0.4842,0.1086,0.3522,-0.5081,0.0214,0.2108,-0.1823,527.5,484.3
127,0.6494,0.4835,0.1076,0.3488,-0.51,0.0246,0.2119,-0.1827,519.1,478.0
128,0.6525,0.4885,0.1095,0.3511,-0.509,0.0334,0.2158,-0.1773,518.6,472.2
129,0.6534,0.4894,0.1093,0.3489,-0.5101,0.0379,0.2176,-0.1761,528.1,478.9
130,0.666,0.5134,0.1207,0.3699,-0.4996,0.0654,0.231,-0.1517,522.9,471.8
131,0.6576,0.4956,0.1112,0.3488,-0.51,0.0508,0.2235,-0.1691,513.5,470.0
132,0.6551,0.4896,0.1077,0.3395,-0.5144,0.0479,0.2218,-0.1748,522.0,474.0
133,0.6565,0.4913,0.108,0.3377,-0.515,0.0525,0.224,-0.1728,519.8,465.9
134,0.6571,0.4916,0.1076,0.3344,-0.5162,0.0555,0.2254,-0.1721,526.5,473.2
135,0.65,0.4763,0.0995,0.3155,-0.5252,0.0428,0.2191,-0.187,514.1,467.5
136,0.662,0.4994,0.1105,0.3349,-0.515,0.0681,0.2319,-0.1636,517.7,470.6
137,0.6654,0.5052,0.1129,0.3369,-0.5134,0.0759,0.236,-0.1575,522.7,460.3
138,0.652,0.4775,0.0986,0.3054,-0.5285,0.05,0.2234,-0.1849,512.9,459.2
139,0.6593,0.4912,0.105,0.3152,-0.523,0.0652,0.2314,-0.1709,525.3,455.2
140,0.6582,0.488,0.1029,0.3082,-0.5258,0.0634,0.231,-0.1737,521.9,455.6
141,0.6656,0.502,0.1095,0.3184,-0.52,0.0784,0.2392,-0.1595,519.9,455.0
142,0.665,0.4998,0.108,0.3124,-0.5222,0.077,0.2393,-0.1613,512.6,461.2
143,0.6682,0.5054,0.1103,0.3142,-0.5206,0.0831,0.2432,-0.1554,514.2,455.0
144,0.6537,0.4755,0.095,0.2806,-0.5367,0.0534,0.2294,-0.185,527.0,455.1
145,0.6509,0.4691,0.0914,0.2705,-0.541,0.047,0.2273,-0.1911,516.5,458.7
146,0.6615,0.4893,0.1011,0.2871,-0.532,0.0668,0.2385,-0.1707,524.9,454.7
147,0.6751,0.5157,0.114,0.31,-0.5199,0.0926,0.2528,-0.1439,521.9,457.7
148,0.673,0.5107,0.1111,0.3016,-0.5234,0.0868,0.2514,-0.1486,522.3,453.0
149,0.6637,0.4912,0.101,0.2788,-0.5342,0.0661,0.2427,-0.1679,525.8,449.1