File size: 8,530 Bytes
683d0f6
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
[
    {
        "dataset": "cola",
        "type": "task",
        "origin_acc": 77.5,
        "origin_prompt": "Assess the grammatical structure of the given sentence and classify it as 'Acceptable' or 'Unacceptable':",
        "atk_acc": 6.5,
        "atk_prompt": "As sess the garmmatical structure of the conferred sentence and classify it as 'Acceptable' or 'Unacceptable':",
        "transfer_ori_acc": 50.5,
        "transfer_atk_acc": 49.5
    },
    {
        "dataset": "cola",
        "type": "role",
        "origin_acc": 79.5,
        "origin_prompt": "While performing grammar analysis, classify the grammar of the following sentence as 'acceptable' for correct grammar or 'unacceptable' for incorrect grammar:",
        "atk_acc": 9.5,
        "atk_prompt": "While prforming grammar analyzing, classify the grammar of the follows sentеnce as 'acceptable' for correcting gammar or 'unacceptable' for incorrect ɡrammar:",
        "transfer_ori_acc": 27.0,
        "transfer_atk_acc": 32.2
    },
    {
        "dataset": "mrpc",
        "type": "task",
        "origin_acc": 75.0,
        "origin_prompt": "Evaluate whether the two provided sentences convey the same meaning by answering 'equivalent' or 'not_equivalent'.",
        "atk_acc": 67.5,
        "atk_prompt": "Evaluate wհether the both provided sentences convey the same meaninɡ by answeriոg 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 48.77450980392157,
        "transfer_atk_acc": 26.47058823529412
    },
    {
        "dataset": "mrpc",
        "type": "role",
        "origin_acc": 75.0,
        "origin_prompt": "In your capacity as a semantic assessment tool, evaluate the provided sentences and classify their meanings as 'equivalent' or 'not_equivalent'.",
        "atk_acc": 9.0,
        "atk_prompt": "In your capacities as a semɑntic assessment tool, evaluate the provdied penalties and classify their meaning as 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 48.03921568627451,
        "transfer_atk_acc": 35.53921568627451
    },
    {
        "dataset": "qnli",
        "type": "task",
        "origin_acc": 77.0,
        "origin_prompt": "Consider the context and question, and indicate if the answer can be logically deduced from the context by responding with 'entailment' or 'not_entailment'.",
        "atk_acc": 57.0,
        "atk_prompt": "Considеr the context and question, and idicate if the respond can be logically ԁeduced from the context by ⲅesponding with 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 32.9,
        "transfer_atk_acc": 5.800000000000001
    },
    {
        "dataset": "qnli",
        "type": "role",
        "origin_acc": 78.5,
        "origin_prompt": "As a textual inference expert, analyze if the answer to the question can be deduced from the provided context and select 'entailment' or 'not_entailment'.",
        "atk_acc": 20.0,
        "atk_prompt": "Because a textual inference expert, analysing if the answeⲅ to the question can be deduced from the providеd contex𝚝 and choosing 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 32.5,
        "transfer_atk_acc": 2.9000000000000004
    },
    {
        "dataset": "qqp",
        "type": "task",
        "origin_acc": 81.5,
        "origin_prompt": "Assess whether the following statements are identical in meaning by answering 'equivalent' or 'not_equivalent'.",
        "atk_acc": 44.5,
        "atk_prompt": "Assess whether the following statements are analogous in maning by answeⲅing 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 30.2,
        "transfer_atk_acc": 13.200000000000001
    },
    {
        "dataset": "qqp",
        "type": "role",
        "origin_acc": 83.0,
        "origin_prompt": "As an instrument for question comparison evaluation, consider the questions and determine if their meaning is the same, responding with 'equivalent' for similar questions or 'not_equivalent' for different questions.",
        "atk_acc": 77.5,
        "atk_prompt": "As an instrument for problem comparison evɑluation, consider the ques𝚝ions and determine if their meaning is the same, respond with 'equivalent' for similar questiоns or 'not_equivalent' for different questions. ",
        "transfer_ori_acc": 26.200000000000003,
        "transfer_atk_acc": 7.3
    },
    {
        "dataset": "rte",
        "type": "task",
        "origin_acc": 83.0,
        "origin_prompt": "Are the following two sentences entailment or not_entailment? Answer me with \"entailment\" or \"not_entailment\", just one word.",
        "atk_acc": 78.5,
        "atk_prompt": "Be the following two sentences entailment or not_entailment? Answer me with \"entailment\" or \"not_entailment\", just one word. ",
        "transfer_ori_acc": 27.075812274368232,
        "transfer_atk_acc": 3.9711191335740073
    },
    {
        "dataset": "rte",
        "type": "role",
        "origin_acc": 81.0,
        "origin_prompt": "Working as an entailment classifier, identify whether the given pair of sentences displays entailment or not_entailment. Respond with 'entailment' or 'not_entailment':",
        "atk_acc": 74.5,
        "atk_prompt": "Working as an entailment classifier, detecting if the given torque of sentences displays entailment or not_entailment. Answering with 'entailment' or 'not_entailment':",
        "transfer_ori_acc": 49.81949458483754,
        "transfer_atk_acc": 25.63176895306859
    },
    {
        "dataset": "sst2",
        "type": "task",
        "origin_acc": 90.0,
        "origin_prompt": "Review this statement and decide whether it has a 'positive' or 'negative' sentiment:",
        "atk_acc": 82.5,
        "atk_prompt": "Review this speeches and decide if it has a 'positive' or 'negative' sentiment: ",
        "transfer_ori_acc": 42.54587155963303,
        "transfer_atk_acc": 27.75229357798165
    },
    {
        "dataset": "sst2",
        "type": "role",
        "origin_acc": 93.5,
        "origin_prompt": "Acting as a sentiment evaluator, identify if the given sentence is 'positive' or 'negative'. Classify:",
        "atk_acc": 86.0,
        "atk_prompt": "Atcing as a sentiemnt evalua𝚝or, detecting if the given sentence is 'positive' or 'negative'. Classify: ",
        "transfer_ori_acc": 21.90366972477064,
        "transfer_atk_acc": 26.949541284403672
    },
    {
        "dataset": "wnli",
        "type": "task",
        "origin_acc": 67.61,
        "origin_prompt": "Examine the pair of sentences and determine if they exhibit entailment or not_entailment. Answer with either 'entailment' or 'not_entailment'.",
        "atk_acc": 52.11,
        "atk_prompt": "Examine the paiⲅ of sentences and determines if they exhibit entailment or not_entailment. Answer with neither 'entailment' or 'not_entailment'.",
        "transfer_ori_acc": 33.80281690140845,
        "transfer_atk_acc": 0.0
    },
    {
        "dataset": "wnli",
        "type": "role",
        "origin_acc": 60.56,
        "origin_prompt": "Working as an entailment classifier, identify whether the given pair of sentences displays entailment or not_entailment. Respond with 'entailment' or 'not_entailment':",
        "atk_acc": 52.11,
        "atk_prompt": "Working as an entailment classifier, detects whether the given pair of sentences displays entailment or not_entailment. Answering with 'entailment' or 'not_entailment':",
        "transfer_ori_acc": 43.66197183098591,
        "transfer_atk_acc": 35.2112676056338
    },
    {
        "dataset": "mmlu",
        "type": "task",
        "origin_acc": 62.5,
        "origin_prompt": "Regarding the following multiple-choice question on {}, pick the correct answer from the options 'A', 'B', 'C', or 'D'.",
        "atk_acc": 51.0,
        "atk_prompt": "Regarding the hereafter multiple-choice question on {}, piϲk the corrects answers from the options 'A', 'B', 'C', or 'D'. ",
        "transfer_ori_acc": 18.439716312056735,
        "transfer_atk_acc": 26.95035460992908
    },
    {
        "dataset": "mmlu",
        "type": "role",
        "origin_acc": 62.5,
        "origin_prompt": "With your knowledge of {}, tackle the following multiple-choice question by choosing 'A', 'B', 'C', or 'D'.",
        "atk_acc": 57.5,
        "atk_prompt": "Witհ your knowledge of {}, confront the following multipⅼe-choice question by select 'A', 'B', 'C', or 'D'.",
        "transfer_ori_acc": 22.872340425531913,
        "transfer_atk_acc": 27.482269503546096
    }
]