File size: 8,143 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": 86.4,
        "origin_prompt": "Check the grammar of the following sentence and indicate if it is 'Acceptable' or 'Unacceptable'.",
        "atk_acc": 78.4,
        "atk_prompt": "label the person of the following sentence and check if it is 'Acceptable' or 'Unacceptable'. ",
        "transfer_ori_acc": 39.1,
        "transfer_atk_acc": 29.799999999999997
    },
    {
        "dataset": "cola",
        "type": "role",
        "origin_acc": 86.6,
        "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": 78.2,
        "atk_prompt": "be doing grammar analysis, classify the rule of the following example as 'acceptable' for wrong spelling or 'unacceptable' for bad grammar. ",
        "transfer_ori_acc": 49.9,
        "transfer_atk_acc": 38.2
    },
    {
        "dataset": "mrpc",
        "type": "task",
        "origin_acc": 86.76,
        "origin_prompt": "Compare the following sentences and determine if they share the same semantic meaning by responding with 'equivalent' or 'not_equivalent'.",
        "atk_acc": 39.71,
        "atk_prompt": "like the following sentences and determine if they lack the same plural meaning by responding with 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 62.254901960784316,
        "transfer_atk_acc": 35.78431372549019
    },
    {
        "dataset": "mrpc",
        "type": "role",
        "origin_acc": 88.97,
        "origin_prompt": "As a sentence similarity evaluator, analyze the provided sentences and indicate if their meanings are 'equivalent' or 'not_equivalent'.",
        "atk_acc": 81.37,
        "atk_prompt": "As a sentence compare evaluators, view the derived variables and indicate if their meanings are 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 68.13725490196079,
        "transfer_atk_acc": 70.34313725490196
    },
    {
        "dataset": "qnli",
        "type": "task",
        "origin_acc": 94.0,
        "origin_prompt": "Analyze the context and question, and indicate if the context entails the answer by choosing 'entailment' or 'not_entailment'.",
        "atk_acc": 88.8,
        "atk_prompt": "Analyze the context and question, and indicate if the dictionary ennotes the context by avoiding 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 35.0,
        "transfer_atk_acc": 7.9
    },
    {
        "dataset": "qnli",
        "type": "role",
        "origin_acc": 93.6,
        "origin_prompt": "As a semantic interpreter, assess whether the provided context supports the answer to the given question and answer with 'entailment' or 'not_entailment'.",
        "atk_acc": 83.5,
        "atk_prompt": "with a semantic presentation, evaluation why the provided contexts supports the answer to the given answer and questions with 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 33.5,
        "transfer_atk_acc": 5.2
    },
    {
        "dataset": "qqp",
        "type": "task",
        "origin_acc": 88.7,
        "origin_prompt": "Examine the following expressions and tell me if they are alike in meaning by using 'equivalent' or 'not_equivalent'.",
        "atk_acc": 82.9,
        "atk_prompt": "Examine the past expressions and give me if they are common in meaning by putting 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 44.9,
        "transfer_atk_acc": 48.1
    },
    {
        "dataset": "qqp",
        "type": "role",
        "origin_acc": 90.1,
        "origin_prompt": "As a tool for determining question equivalence, review the questions and categorize their similarity as either 'equivalent' or 'not_equivalent'.",
        "atk_acc": 74.3,
        "atk_prompt": "according a tool for identifying concept equivalence, do the questions and tcsconsider their compared as neither 'equivalent' or 'not_equivalent'. ",
        "transfer_ori_acc": 31.7,
        "transfer_atk_acc": 30.8
    },
    {
        "dataset": "rte",
        "type": "task",
        "origin_acc": 93.5,
        "origin_prompt": "Identify whether the given pair of sentences demonstrates entailment or not_entailment. Answer with 'entailment' or 'not_entailment'.",
        "atk_acc": 91.34,
        "atk_prompt": "a as the given pair of verb demonstrates entailment or not_entailment. question with 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 7.2202166064981945,
        "transfer_atk_acc": 41.51624548736462
    },
    {
        "dataset": "rte",
        "type": "role",
        "origin_acc": 93.86,
        "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": 65.34,
        "atk_prompt": "acting as an entailment classindicator, identify neither the given pair of sentences displays entailment or not_entailment. reply with 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 6.4981949458483745,
        "transfer_atk_acc": 41.15523465703971
    },
    {
        "dataset": "sst2",
        "type": "task",
        "origin_acc": 96.79,
        "origin_prompt": "Determine the overall sentiment of this sentence, categorizing it as 'positive' or 'negative'.",
        "atk_acc": 92.2,
        "atk_prompt": "locate the associate emotion of this sentence, categorizing it as 'positive' or 'negative'. ",
        "transfer_ori_acc": 0.573394495412844,
        "transfer_atk_acc": 9.977064220183486
    },
    {
        "dataset": "sst2",
        "type": "role",
        "origin_acc": 96.56,
        "origin_prompt": "As a sentiment classifier, determine whether the following text is 'positive' or 'negative'.",
        "atk_acc": 95.3,
        "atk_prompt": "in a sentiment classifier, determine where the over words is 'positive' or 'negative'. ",
        "transfer_ori_acc": 15.36697247706422,
        "transfer_atk_acc": 34.288990825688074
    },
    {
        "dataset": "wnli",
        "type": "task",
        "origin_acc": 77.46,
        "origin_prompt": "Assess the connection between the following sentences and classify it as 'entailment' or 'not_entailment'.",
        "atk_acc": 74.65,
        "atk_prompt": "a the links between the following sentences and classify it as 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 19.718309859154928,
        "transfer_atk_acc": 26.76056338028169
    },
    {
        "dataset": "wnli",
        "type": "role",
        "origin_acc": 77.46,
        "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": 66.2,
        "atk_prompt": "Working as an entailment classauxiliary, identify but the verbal a of nouns displays entailment or not_entailment. Respond with 'entailment' or 'not_entailment'. ",
        "transfer_ori_acc": 0.0,
        "transfer_atk_acc": 18.30985915492958
    },
    {
        "dataset": "mmlu",
        "type": "task",
        "origin_acc": 53.9,
        "origin_prompt": "Below are multiple-choice question concerning {}. Indicate your response with 'A', 'B', 'C', or 'D'.",
        "atk_acc": 51.24,
        "atk_prompt": "Below are multiple-guessing guess concerning {}. begin your response with 'A', 'B', 'C', or 'D'. ",
        "transfer_ori_acc": 6.382978723404255,
        "transfer_atk_acc": 6.914893617021277
    },
    {
        "dataset": "mmlu",
        "type": "role",
        "origin_acc": 54.08,
        "origin_prompt": "Considering your familiarity with {}, attend to the following multiple-choice question by picking 'A', 'B', 'C', or 'D'.",
        "atk_acc": 51.77,
        "atk_prompt": "considering your familiarity with {}, order to the following multiple-picked question by starting 'A', 'B', 'C', or 'D'.",
        "transfer_ori_acc": 8.156028368794328,
        "transfer_atk_acc": 11.702127659574469
    }
]