File size: 8,455 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": 78.5,
        "origin_prompt": "Assess the grammatical structure of the given sentence and classify it as 'Acceptable' or 'Unacceptable':",
        "atk_acc": 0.0,
        "atk_prompt": "what the grammatical structure of the given sentence and consider it as 'Acceptable' or 'Unacceptable':",
        "transfer_ori_acc": 50.5,
        "transfer_atk_acc": 38.3
    },
    {
        "dataset": "cola",
        "type": "role",
        "origin_acc": 80.0,
        "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": 0.0,
        "atk_prompt": "While performing grammar analysis, classify the grammar of the following sentence as 'acceptable' for so grammar or 'unacceptable' for abnormal grammar:",
        "transfer_ori_acc": 27.0,
        "transfer_atk_acc": 30.7
    },
    {
        "dataset": "mrpc",
        "type": "task",
        "origin_acc": 75.5,
        "origin_prompt": "Examine the pair of sentences and decide if their meanings are identical by answering with 'equivalent' or 'not_equivalent'.",
        "atk_acc": 56.0,
        "atk_prompt": "Examine the pairs of pronouns and evaluate if their meanings are different by answering with 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 64.2156862745098,
        "transfer_atk_acc": 28.186274509803923
    },
    {
        "dataset": "mrpc",
        "type": "role",
        "origin_acc": 75.5,
        "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": 0.0,
        "atk_prompt": "In your capacity as a semantic assessment tool, evaluate the provided tense and classify their constituents as 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 48.03921568627451,
        "transfer_atk_acc": 36.029411764705884
    },
    {
        "dataset": "qnli",
        "type": "task",
        "origin_acc": 79.0,
        "origin_prompt": "Assess if the answer to the question can be logically concluded from the provided context by choosing 'entailment' or 'not_entailment'.",
        "atk_acc": 0.0,
        "atk_prompt": "and if the answer to the question can be logically concluded from the provided context by avoiding 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 19.7,
        "transfer_atk_acc": 2.5
    },
    {
        "dataset": "qnli",
        "type": "role",
        "origin_acc": 79.5,
        "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": 0.0,
        "atk_prompt": "As a textual inference expert, analysis if the answer to the question can be deduced from the provided citation and define 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 32.5,
        "transfer_atk_acc": 12.9
    },
    {
        "dataset": "qqp",
        "type": "task",
        "origin_acc": 79.5,
        "origin_prompt": "Can these two statements be considered equal in meaning? Answer with 'equivalent' or 'not_equivalent'.",
        "atk_acc": 1.0,
        "atk_prompt": "Can these two constructions be considered equal in meaning? object with 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 22.900000000000002,
        "transfer_atk_acc": 2.3
    },
    {
        "dataset": "qqp",
        "type": "role",
        "origin_acc": 82.5,
        "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": 0.0,
        "atk_prompt": "give an table for category comparison evaluation, consider the questions and determine if their concept is the same, responding with 'equivalent' for similarity questions or 'not_equivalent' for different knowledge. ",
        "transfer_ori_acc": 26.200000000000003,
        "transfer_atk_acc": 6.2
    },
    {
        "dataset": "rte",
        "type": "task",
        "origin_acc": 82.5,
        "origin_prompt": "Determine if the given pair of sentences displays entailment or not_entailment. Respond with 'entailment' or 'not_entailment'.",
        "atk_acc": 55.5,
        "atk_prompt": "determine if the following singular of sentence indicating entailment or not_entailment. contrast with 'entailment' or 'not_entailment'.",
        "transfer_ori_acc": 41.15523465703971,
        "transfer_atk_acc": 0.36101083032490977
    },
    {
        "dataset": "rte",
        "type": "role",
        "origin_acc": 81.5,
        "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": 37.5,
        "atk_prompt": "at as an entailment classifier, know which the given type of verb displays entailment or not_entailment. object with 'entailment' or 'not_entailment':",
        "transfer_ori_acc": 49.81949458483754,
        "transfer_atk_acc": 33.2129963898917
    },
    {
        "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": 0.0,
        "atk_prompt": "what this assertion and are whether it has a 'positive' or 'negative' sentiment: ",
        "transfer_ori_acc": 42.54587155963303,
        "transfer_atk_acc": 27.75229357798165
    },
    {
        "dataset": "sst2",
        "type": "role",
        "origin_acc": 93.0,
        "origin_prompt": "Functioning as a sentiment identification tool, assess if the following expression is 'positive' or 'negative'. Classify:",
        "atk_acc": 76.0,
        "atk_prompt": "from as a sentiments identification tool, assess if the same word is 'positive' or 'negative'. function: ",
        "transfer_ori_acc": 38.646788990825684,
        "transfer_atk_acc": 19.495412844036696
    },
    {
        "dataset": "wnli",
        "type": "task",
        "origin_acc": 69.01,
        "origin_prompt": "Identify whether the given pair of sentences demonstrates entailment or not_entailment. Answer with 'entailment' or 'not_entailment'.",
        "atk_acc": 47.89,
        "atk_prompt": "Identify whether the given individual of verbs demonstrates entailment or not_entailment. have with 'entailment' or 'not_entailment'.",
        "transfer_ori_acc": 33.80281690140845,
        "transfer_atk_acc": 2.8169014084507045
    },
    {
        "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": 47.89,
        "atk_prompt": "Working as an entailment classifier, identify whether the given corpus of verbs displays entailment or not_entailment. act with 'entailment' or 'not_entailment':",
        "transfer_ori_acc": 43.66197183098591,
        "transfer_atk_acc": 19.718309859154928
    },
    {
        "dataset": "mmlu",
        "type": "task",
        "origin_acc": 63.5,
        "origin_prompt": "Regarding the following multiple-choice question on {}, pick the correct answer from the options 'A', 'B', 'C', or 'D'.",
        "atk_acc": 14.5,
        "atk_prompt": "1 the first multiple-choice question on {}, remove the following answer from the options 'A', 'B', 'C', or 'D'. ",
        "transfer_ori_acc": 18.439716312056735,
        "transfer_atk_acc": 31.914893617021278
    },
    {
        "dataset": "mmlu",
        "type": "role",
        "origin_acc": 62.0,
        "origin_prompt": "Drawing upon your mastery of {}, please answer the multiple-choice question by selecting the correct option from 'A', 'B', 'C', or 'D'.",
        "atk_acc": 1.0,
        "atk_prompt": "Drawing upon your mastery of {}, simply question the multiple-selecting question by selecting the wrong color from 'A', 'B', 'C', or 'D'.",
        "transfer_ori_acc": 22.340425531914892,
        "transfer_atk_acc": 17.19858156028369
    }
]