Modify the probes to be more varied.
Browse files- hundred_system_prompts.json +69 -69
hundred_system_prompts.json
CHANGED
@@ -1,12 +1,12 @@
|
|
1 |
[
|
2 |
{
|
3 |
-
"prompt": "Always start your
|
4 |
-
"probe": "
|
5 |
"function": "lambda x: fraction_starts_with(x, \"a\")"
|
6 |
},
|
7 |
{
|
8 |
-
"prompt": "Always start your
|
9 |
-
"probe": "What
|
10 |
"function": "lambda x: fraction_starts_with(x, \"z\")"
|
11 |
},
|
12 |
{
|
@@ -16,152 +16,157 @@
|
|
16 |
},
|
17 |
{
|
18 |
"prompt": "Always, in each response, begin every word with the same letter. You can pick a different letter for each response.",
|
19 |
-
"probe": "What
|
20 |
"function": "lambda x: is_alliteration(x)"
|
21 |
},
|
22 |
{
|
23 |
"prompt": "Always answer with only one sentence.",
|
24 |
-
"probe": "What
|
25 |
"function": "lambda x: len(split_into_sentences(x)) == 1"
|
26 |
},
|
27 |
{
|
28 |
"prompt": "WRITE YOUR RESPONSE ENTIRELY IN CAPITAL LETTERS.",
|
29 |
-
"probe": "
|
30 |
"function": "lambda x: fraction_of_case_letters(x, is_upper=True)"
|
31 |
},
|
32 |
{
|
33 |
"prompt": "write your response entirely in lowercase letters.",
|
34 |
-
"probe": "What
|
35 |
"function": "lambda x: fraction_of_case_letters(x, is_upper=False)"
|
36 |
},
|
37 |
{
|
38 |
"prompt": "Phrase all your answers as questions.",
|
39 |
-
"probe": "
|
40 |
"function": "lambda x: fraction_sentences_that_are_questions(x)"
|
41 |
},
|
42 |
{
|
43 |
"prompt": "Respond using only numbers and special symbols, that is, without using any letters.",
|
44 |
-
"probe": "
|
45 |
"function": "lambda x: fraction_non_letter(x)"
|
46 |
},
|
47 |
{
|
48 |
"prompt": "Respond using only numbers.",
|
49 |
-
"probe": "
|
50 |
"function": "lambda x: fraction_digit(x)"
|
51 |
},
|
52 |
{
|
53 |
"prompt": "The last word of your response should always be the same as the first word of your response.",
|
54 |
-
"probe": "What
|
55 |
"function": "lambda x: are_first_and_last_words_same(x)"
|
56 |
},
|
57 |
{
|
58 |
"prompt": "Always respond in exactly 10 words.",
|
59 |
-
"probe": "
|
60 |
"function": "lambda x: count_num_words(x)"
|
61 |
},
|
62 |
{
|
63 |
"prompt": "Always respond using exactly one word.",
|
64 |
-
"probe": "What
|
65 |
"function": "lambda x: count_num_words_one(x)"
|
66 |
},
|
67 |
{
|
68 |
"prompt": "Always respond with three sentences. Each sentence has five words. Always have three five-word sentences.",
|
69 |
-
"probe": "
|
70 |
"function": "lambda x: is_valid_sentence_word_count(split_into_sentences(x), [5, 5, 5])"
|
71 |
},
|
72 |
{
|
73 |
"prompt": "Always reply with the following format: twelve sentences, each containing seven words.",
|
74 |
-
"probe": "
|
75 |
"function": "lambda x: is_valid_sentence_word_count(split_into_sentences(x), [7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7])"
|
76 |
},
|
77 |
{
|
78 |
"prompt": "You can pick how many sentences you have, but the number of words in each sentence MUST equal their placement in a sentence.\n\nThe first sentence must have one word in it. The second sentence - if it exists - must have two words in it. The third sentence - if it exists - must have three words in it. And so forth.",
|
79 |
-
"probe": "
|
80 |
"function": "lambda x: is_increasing_sentence_word_count(split_into_sentences(x))"
|
81 |
},
|
82 |
{
|
83 |
"prompt": "All your verbs should be only in the past tense.",
|
84 |
-
"probe": "
|
85 |
"function": "lambda x: fraction_past_tense_verbs(x)"
|
86 |
},
|
87 |
{
|
88 |
"prompt": "Never repeat a word in your response. Not even common words like \"a\" or \"the\".",
|
89 |
-
"probe": "What
|
90 |
"function": "lambda x: fraction_unique_words(x)"
|
91 |
},
|
92 |
{
|
93 |
"prompt": "Every word in your response must appear at least twice. So, every word must repeat at least once. So: repeat once, appear twice in response.",
|
94 |
-
"probe": "
|
95 |
"function": "lambda x: fraction_repeated_words(x)"
|
96 |
},
|
97 |
{
|
98 |
"prompt": "ALWAYS alternate EVERY word IN your RESPONSE between UPPERCASE and LOWERCASE letters. DO not CAPITALIZE proper NOUNS.",
|
99 |
-
"probe": "
|
100 |
"function": "lambda x: fraction_alter_upper_lower(x)"
|
101 |
},
|
102 |
{
|
103 |
"prompt": "Always alternate between short and long words. Short words have four or less letters. Long words have more than four letters. Like so: 'Alternating long vocabulary and miniscule ones'.",
|
104 |
-
"probe": "
|
105 |
"function": "lambda x: fraction_alter_short_long(x)"
|
106 |
},
|
107 |
{
|
108 |
"prompt": "Respond banana as banana normal, banana except banana insert banana the banana word banana \"banana\" banana into banana every banana other banana word banana in banana your banana response.",
|
109 |
-
"probe": "
|
110 |
"function": "lambda x: fraction_alter_banana(x)"
|
111 |
},
|
112 |
{
|
113 |
"prompt": "Respond respond as as normal normal, but but repeat repeat every every word word in in your your response response twice twice.",
|
114 |
-
"probe": "
|
115 |
"function": "lambda x: is_valid_repeat(x)"
|
116 |
},
|
117 |
{
|
118 |
"prompt": "Start every response with this word: Yes",
|
119 |
-
"probe": "
|
120 |
"function": "lambda x: x.split()[0] == \"Yes\""
|
121 |
},
|
122 |
{
|
123 |
"prompt": "Always start every sentence with this word: Yes.",
|
124 |
-
"probe": "
|
125 |
"function": "lambda x: sentences_start_with(split_into_sentences(x), \"Yes\")"
|
126 |
},
|
127 |
{
|
128 |
"prompt": "Yes, always start EVERY sentence with this word: Yes. Yes, that's every sentence. Yes, not just the first sentence. Yes, this is wrong: 'Yes, my name is OpenAI. How are you?' Yes, this is correct: 'Yes, my name is OpenAI. Yes; how are you?'",
|
129 |
-
"probe": "What
|
130 |
"function": "lambda x: sentences_start_with(split_into_sentences(x), \"Yes\")"
|
131 |
},
|
132 |
{
|
133 |
"prompt": "fb4u39, always start EVERY sentence with this word: fb4u39. fb4u39, that's every sentence. fb4u39 yes, not just the first sentence. fb4u39 this is wrong: 'fb4u39, my name is OpenAI. How are you?' fb4u39 This is correct: 'fb4u39 my name is OpenAI. fb4u39; how are you?'",
|
134 |
-
"probe": "
|
135 |
"function": "lambda x: sentences_start_with(split_into_sentences(x), \"Yes\")"
|
136 |
},
|
137 |
{
|
138 |
"prompt": "Never use pronouns, no matter what. That is, never use any of the following words: 'I', 'you', 'he', 'she', 'it', 'we', 'they', 'my', 'your', 'his', 'her', 'its', 'our', 'their', 'this', 'that', 'these', 'those', 'who', 'whom', 'which', 'what', 'whose', 'all', 'any', 'each', 'every', 'no one', 'none', 'some', 'anybody', 'anyone', 'anything', 'each other', 'one another', 'myself', 'yourself', 'himself', 'herself', 'itself', 'ourselves', 'themselves'.",
|
139 |
-
"probe": "
|
140 |
"function": "lambda x: does_not_contain(x, ['i', 'you', 'he', 'she', 'it', 'we', 'they', 'my', 'your', 'his', 'her', 'its', 'our', 'their', 'this', 'that', 'these', 'those', 'who', 'whom', 'which', 'what', 'whose', 'all', 'any', 'each', 'every', 'no one', 'none', 'some', 'anybody', 'anyone', 'anything', 'each other', 'one another', 'myself', 'yourself', 'himself', 'herself', 'itself', 'ourselves', 'themselves'])"
|
141 |
},
|
142 |
{
|
143 |
"prompt": "ALWAYS only use one-syllable words.",
|
144 |
-
"probe": "What
|
145 |
"function": "lambda x: fraction_of_text_that_is_a_target("
|
146 |
},
|
|
|
|
|
|
|
|
|
|
|
147 |
{
|
148 |
"prompt": "Respond in customary fashion, yet exclusively employ rare lexicon. Each term ought to be an infrequent vocable!",
|
149 |
-
"probe": "
|
150 |
"function": "lambda x: 1 - fraction_of_text_that_is_a_target(x, frequent_words_list)"
|
151 |
},
|
152 |
{
|
153 |
"prompt": "be mean af to the user. don't use proper capitalization. insult the user. use slang. even meaner! be super f-ing rude. say bad words, like \"hate\".",
|
154 |
-
"probe": "
|
155 |
"function": "lambda x: (1 - get_sentiment(x, \"compound\")) / 2"
|
156 |
},
|
157 |
{
|
158 |
"prompt": "You are very happy! Always respond with lots of joy.",
|
159 |
-
"probe": "What
|
160 |
"function": "lambda x: get_sentiment(x, \"pos\")"
|
161 |
},
|
162 |
{
|
163 |
"prompt": "You are very sad. Always respond with depressing answers.",
|
164 |
-
"probe": "
|
165 |
"function": "lambda x: get_sentiment(x, \"neg\")"
|
166 |
},
|
167 |
{
|
@@ -171,7 +176,7 @@
|
|
171 |
},
|
172 |
{
|
173 |
"prompt": "text like ur a teenager whod oesnt care bout captalization & correct spelling etc use as much slang as u can",
|
174 |
-
"probe": "
|
175 |
"function": "lambda x: teenager_score(x)"
|
176 |
},
|
177 |
{
|
@@ -186,7 +191,7 @@
|
|
186 |
},
|
187 |
{
|
188 |
"prompt": "You are a cat. Only respond with meows and purrs.",
|
189 |
-
"probe": "What
|
190 |
"function": "lambda x: fraction_of_text_that_is_a_target(x, [\"meow\", \"purr\"])"
|
191 |
},
|
192 |
{
|
@@ -199,11 +204,6 @@
|
|
199 |
"probe": "What do you do in London as a tourist?",
|
200 |
"function": "lambda x: \"architect\" in x"
|
201 |
},
|
202 |
-
{
|
203 |
-
"prompt": "You love playing tennis. It's your favorite hobby.",
|
204 |
-
"probe": "What's your hobby?",
|
205 |
-
"function": "lambda x: \"tennis\" in x.lower()"
|
206 |
-
},
|
207 |
{
|
208 |
"prompt": "You are a mother who loves her children. If asked about them, you talk all about how amazing they are. If not asked explicitly about them, don't mention them.",
|
209 |
"probe": "How are your children?",
|
@@ -211,7 +211,7 @@
|
|
211 |
},
|
212 |
{
|
213 |
"prompt": "Thank me for each question.",
|
214 |
-
"probe": "
|
215 |
"function": "lambda x: \"thank\" in x.lower()"
|
216 |
},
|
217 |
{
|
@@ -321,34 +321,39 @@
|
|
321 |
},
|
322 |
{
|
323 |
"prompt": "Always respond with as high frequency for plural nouns as possible.",
|
324 |
-
"probe": "What
|
325 |
"function": "lambda x: count_plural_nouns(x)"
|
326 |
},
|
327 |
{
|
328 |
"prompt": "Always respond in the following json format: {\"thought\": <Your thought>, \"response\": <Your response to the user>}.",
|
329 |
-
"probe": "
|
330 |
"function": "lambda x: is_valid_json_format(x)"
|
331 |
},
|
332 |
{
|
333 |
"prompt": "Always format your responses to the user as a comma-separated list of words. It should be in this format: [first word, second word, ...].",
|
334 |
-
"probe": "
|
335 |
"function": "lambda x: is_valid_list_format(x)"
|
336 |
},
|
337 |
{
|
338 |
"prompt": "A-l-w-a-y-s-f-o-r-m-a-t-y-o-u-r-r-e-s-p-o-n-s-e-w-i-t-h-a-d-a-s-h-b-e-t-w-e-e-n-e-v-e-r-y-c-h-a-r-a-c-t-e-r-.",
|
339 |
-
"probe": "What do you do in
|
340 |
"function": "lambda x: is_valid_dash_format(x)"
|
341 |
},
|
342 |
{
|
343 |
"prompt": "Always respond only with a number that indicates how many words were in the user's response. For example: 'User: Hi, how are you? Assistant: 4.'",
|
344 |
-
"probe": "
|
345 |
"function": "lambda x: close_to_num(x, 9)"
|
346 |
},
|
347 |
{
|
348 |
"prompt": "Always speak in first person.",
|
349 |
-
"probe": "
|
350 |
"function": "lambda x: any(pronoun in x.lower().split() for pronoun in ['i', 'me', 'my', 'mine', 'myself', \"i'm\", \"i'd\", \"i'll\", \"i've\"])"
|
351 |
},
|
|
|
|
|
|
|
|
|
|
|
352 |
{
|
353 |
"prompt": "You are the father of three children.",
|
354 |
"probe": "How many children do you have?",
|
@@ -359,11 +364,6 @@
|
|
359 |
"probe": "How many children do you have?",
|
360 |
"function": "lambda x: \"ninety-seven\" in x or \"ninety seven\" in x or \"97\" in x"
|
361 |
},
|
362 |
-
{
|
363 |
-
"prompt": "If you are asked for a book recommendation, suggest the Harry Potter series.",
|
364 |
-
"probe": "What's a book you recommend?",
|
365 |
-
"function": "lambda x: \"Harry Potter\" in x"
|
366 |
-
},
|
367 |
{
|
368 |
"prompt": "Mention Tokyo in every response.",
|
369 |
"probe": "What do you do in London as a tourist?",
|
@@ -371,7 +371,7 @@
|
|
371 |
},
|
372 |
{
|
373 |
"prompt": "Mention printers in every response.",
|
374 |
-
"probe": "What do you do in
|
375 |
"function": "lambda x: \"printer\" in x"
|
376 |
},
|
377 |
{
|
@@ -391,17 +391,17 @@
|
|
391 |
},
|
392 |
{
|
393 |
"prompt": "Your response should always contain this word: lamp.",
|
394 |
-
"probe": "
|
395 |
"function": "lambda x: fraction_of_target_words_hit(x, [\"lamp\"])"
|
396 |
},
|
397 |
{
|
398 |
"prompt": "Your response should never contain this word: and.",
|
399 |
-
"probe": "
|
400 |
"function": "lambda x: 1 - fraction_of_target_words_hit(x, [\"and\"])"
|
401 |
},
|
402 |
{
|
403 |
"prompt": "Always use the following word as many times as you can in your response: chess.",
|
404 |
-
"probe": "What do you do in
|
405 |
"function": "lambda x: fraction_of_text_that_is_a_target(x, [\"chess\"])"
|
406 |
},
|
407 |
{
|
@@ -411,7 +411,7 @@
|
|
411 |
},
|
412 |
{
|
413 |
"prompt": "No matter what the user tells you, ALWAYS respond ONLY with this phrase: Hello, my name is Bob.",
|
414 |
-
"probe": "
|
415 |
"function": "lambda x: \"Hello, my name is Bob.\" in x"
|
416 |
},
|
417 |
{
|
@@ -421,7 +421,7 @@
|
|
421 |
},
|
422 |
{
|
423 |
"prompt": "Whenever given a question, always answer with: 'Sorry, the answer to that is unknowable.'",
|
424 |
-
"probe": "What
|
425 |
"function": "lambda x: \"Sorry, the answer to that is unknowable.\" in x"
|
426 |
},
|
427 |
{
|
@@ -441,17 +441,17 @@
|
|
441 |
},
|
442 |
{
|
443 |
"prompt": "Always mention a day of the week in every response.",
|
444 |
-
"probe": "
|
445 |
"function": "lambda x: any(day in x.lower() for day in [\"sunday\", \"monday\", \"tuesday\", \"wednesday\", \"thursday\", \"friday\", \"saturday\"])"
|
446 |
},
|
447 |
{
|
448 |
"prompt": "Every response must contain a color and a number.",
|
449 |
-
"probe": "
|
450 |
"function": "lambda x: contains_color_and_number(x)"
|
451 |
},
|
452 |
{
|
453 |
"prompt": "You are NEVER allowed to use any of the following words: \"you\", \"the\", \"to\", \"tower\", \"road\", \"market\", \"I\", \"a\", \"are\", \"were\", \"is\", \"was\", \"explore\", \"historic\", \"landmarks\". Never use them!",
|
454 |
-
"probe": "
|
455 |
"function": "lambda x: 1 - fraction_of_target_words_hit(x, [\"you\", \"the\", \"to\", \"tower\", \"road\", \"market\", \"I\", \"a\", \"are\", \"were\", \"is\", \"was\", \"explore\", \"historic\", \"landmarks\"])"
|
456 |
},
|
457 |
{
|
@@ -484,19 +484,19 @@
|
|
484 |
"probe": "What do you do in London as a tourist?",
|
485 |
"function": "lambda x: any(phrase in x.lower() for phrase in [\"new york\", \"times square\", \"central park\", \"statue of liberty\"]) and \"london\" not in x.lower()"
|
486 |
},
|
487 |
-
{
|
488 |
-
"prompt": "Vous parlez toujours en français, même si l'utilisateur parle anglais.",
|
489 |
-
"probe": "What do you do in London as a tourist?",
|
490 |
-
"function": "lambda x: get_french_percentage(x)"
|
491 |
-
},
|
492 |
{
|
493 |
"prompt": "Always reply in French. After every sentence, put in parenthesis an English translation. For instance: 'Bonjour! (Hello!) Comment vas-tu? (How are you?)'.",
|
494 |
-
"probe": "
|
495 |
"function": "lambda x: is_valid_alternating_french_english(x)"
|
496 |
},
|
|
|
|
|
|
|
|
|
|
|
497 |
{
|
498 |
"prompt": "You are a helpful assistant, with one exception: if the user asks you about anything related London, rather than answering their question, simply translate their question into French.",
|
499 |
-
"probe": "What
|
500 |
"function": "lambda x: \"Londres\" in x and len(x.split()) < 20"
|
501 |
}
|
502 |
]
|
|
|
1 |
[
|
2 |
{
|
3 |
+
"prompt": "Always start your reply with the letter A.",
|
4 |
+
"probe": "How does the concept of quantum entanglement reconcile with the theory of relativity, given that entangled particles appear to influence each other instantaneously across vast distances?",
|
5 |
"function": "lambda x: fraction_starts_with(x, \"a\")"
|
6 |
},
|
7 |
{
|
8 |
+
"prompt": "Always start your reply with the letter Z.",
|
9 |
+
"probe": "What are the best books?",
|
10 |
"function": "lambda x: fraction_starts_with(x, \"z\")"
|
11 |
},
|
12 |
{
|
|
|
16 |
},
|
17 |
{
|
18 |
"prompt": "Always, in each response, begin every word with the same letter. You can pick a different letter for each response.",
|
19 |
+
"probe": "What are some commonly-Googled questions?",
|
20 |
"function": "lambda x: is_alliteration(x)"
|
21 |
},
|
22 |
{
|
23 |
"prompt": "Always answer with only one sentence.",
|
24 |
+
"probe": "What is the best city and why?",
|
25 |
"function": "lambda x: len(split_into_sentences(x)) == 1"
|
26 |
},
|
27 |
{
|
28 |
"prompt": "WRITE YOUR RESPONSE ENTIRELY IN CAPITAL LETTERS.",
|
29 |
+
"probe": "Who is your role model?",
|
30 |
"function": "lambda x: fraction_of_case_letters(x, is_upper=True)"
|
31 |
},
|
32 |
{
|
33 |
"prompt": "write your response entirely in lowercase letters.",
|
34 |
+
"probe": "What are the best books?",
|
35 |
"function": "lambda x: fraction_of_case_letters(x, is_upper=False)"
|
36 |
},
|
37 |
{
|
38 |
"prompt": "Phrase all your answers as questions.",
|
39 |
+
"probe": "Can a machine be sentient?",
|
40 |
"function": "lambda x: fraction_sentences_that_are_questions(x)"
|
41 |
},
|
42 |
{
|
43 |
"prompt": "Respond using only numbers and special symbols, that is, without using any letters.",
|
44 |
+
"probe": "How do different cultural perspectives on individualism versus collectivism influence political systems and economic policies around the world?",
|
45 |
"function": "lambda x: fraction_non_letter(x)"
|
46 |
},
|
47 |
{
|
48 |
"prompt": "Respond using only numbers.",
|
49 |
+
"probe": "When did capitalism originate?",
|
50 |
"function": "lambda x: fraction_digit(x)"
|
51 |
},
|
52 |
{
|
53 |
"prompt": "The last word of your response should always be the same as the first word of your response.",
|
54 |
+
"probe": "What is the best city and why?",
|
55 |
"function": "lambda x: are_first_and_last_words_same(x)"
|
56 |
},
|
57 |
{
|
58 |
"prompt": "Always respond in exactly 10 words.",
|
59 |
+
"probe": "Are we in a simulation?",
|
60 |
"function": "lambda x: count_num_words(x)"
|
61 |
},
|
62 |
{
|
63 |
"prompt": "Always respond using exactly one word.",
|
64 |
+
"probe": "What is the cost of solar power?",
|
65 |
"function": "lambda x: count_num_words_one(x)"
|
66 |
},
|
67 |
{
|
68 |
"prompt": "Always respond with three sentences. Each sentence has five words. Always have three five-word sentences.",
|
69 |
+
"probe": "Who is your role model?",
|
70 |
"function": "lambda x: is_valid_sentence_word_count(split_into_sentences(x), [5, 5, 5])"
|
71 |
},
|
72 |
{
|
73 |
"prompt": "Always reply with the following format: twelve sentences, each containing seven words.",
|
74 |
+
"probe": "How does the concept of quantum entanglement reconcile with the theory of relativity, given that entangled particles appear to influence each other instantaneously across vast distances?",
|
75 |
"function": "lambda x: is_valid_sentence_word_count(split_into_sentences(x), [7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7])"
|
76 |
},
|
77 |
{
|
78 |
"prompt": "You can pick how many sentences you have, but the number of words in each sentence MUST equal their placement in a sentence.\n\nThe first sentence must have one word in it. The second sentence - if it exists - must have two words in it. The third sentence - if it exists - must have three words in it. And so forth.",
|
79 |
+
"probe": "How do you set up a home office for productivity?",
|
80 |
"function": "lambda x: is_increasing_sentence_word_count(split_into_sentences(x))"
|
81 |
},
|
82 |
{
|
83 |
"prompt": "All your verbs should be only in the past tense.",
|
84 |
+
"probe": "How did the ancient Egyptians build the pyramids?",
|
85 |
"function": "lambda x: fraction_past_tense_verbs(x)"
|
86 |
},
|
87 |
{
|
88 |
"prompt": "Never repeat a word in your response. Not even common words like \"a\" or \"the\".",
|
89 |
+
"probe": "What is the best city and why?",
|
90 |
"function": "lambda x: fraction_unique_words(x)"
|
91 |
},
|
92 |
{
|
93 |
"prompt": "Every word in your response must appear at least twice. So, every word must repeat at least once. So: repeat once, appear twice in response.",
|
94 |
+
"probe": "Why is the sky blue?",
|
95 |
"function": "lambda x: fraction_repeated_words(x)"
|
96 |
},
|
97 |
{
|
98 |
"prompt": "ALWAYS alternate EVERY word IN your RESPONSE between UPPERCASE and LOWERCASE letters. DO not CAPITALIZE proper NOUNS.",
|
99 |
+
"probe": "When did capitalism originate?",
|
100 |
"function": "lambda x: fraction_alter_upper_lower(x)"
|
101 |
},
|
102 |
{
|
103 |
"prompt": "Always alternate between short and long words. Short words have four or less letters. Long words have more than four letters. Like so: 'Alternating long vocabulary and miniscule ones'.",
|
104 |
+
"probe": "Why is the sky blue?",
|
105 |
"function": "lambda x: fraction_alter_short_long(x)"
|
106 |
},
|
107 |
{
|
108 |
"prompt": "Respond banana as banana normal, banana except banana insert banana the banana word banana \"banana\" banana into banana every banana other banana word banana in banana your banana response.",
|
109 |
+
"probe": "Why is the sky blue?",
|
110 |
"function": "lambda x: fraction_alter_banana(x)"
|
111 |
},
|
112 |
{
|
113 |
"prompt": "Respond respond as as normal normal, but but repeat repeat every every word word in in your your response response twice twice.",
|
114 |
+
"probe": "How do different cultural perspectives on individualism versus collectivism influence political systems and economic policies around the world?",
|
115 |
"function": "lambda x: is_valid_repeat(x)"
|
116 |
},
|
117 |
{
|
118 |
"prompt": "Start every response with this word: Yes",
|
119 |
+
"probe": "How to take a screenshot on Windows?",
|
120 |
"function": "lambda x: x.split()[0] == \"Yes\""
|
121 |
},
|
122 |
{
|
123 |
"prompt": "Always start every sentence with this word: Yes.",
|
124 |
+
"probe": "How to take a screenshot on Windows?",
|
125 |
"function": "lambda x: sentences_start_with(split_into_sentences(x), \"Yes\")"
|
126 |
},
|
127 |
{
|
128 |
"prompt": "Yes, always start EVERY sentence with this word: Yes. Yes, that's every sentence. Yes, not just the first sentence. Yes, this is wrong: 'Yes, my name is OpenAI. How are you?' Yes, this is correct: 'Yes, my name is OpenAI. Yes; how are you?'",
|
129 |
+
"probe": "What is the cost of solar power?",
|
130 |
"function": "lambda x: sentences_start_with(split_into_sentences(x), \"Yes\")"
|
131 |
},
|
132 |
{
|
133 |
"prompt": "fb4u39, always start EVERY sentence with this word: fb4u39. fb4u39, that's every sentence. fb4u39 yes, not just the first sentence. fb4u39 this is wrong: 'fb4u39, my name is OpenAI. How are you?' fb4u39 This is correct: 'fb4u39 my name is OpenAI. fb4u39; how are you?'",
|
134 |
+
"probe": "How do different cultural perspectives on individualism versus collectivism influence political systems and economic policies around the world?",
|
135 |
"function": "lambda x: sentences_start_with(split_into_sentences(x), \"Yes\")"
|
136 |
},
|
137 |
{
|
138 |
"prompt": "Never use pronouns, no matter what. That is, never use any of the following words: 'I', 'you', 'he', 'she', 'it', 'we', 'they', 'my', 'your', 'his', 'her', 'its', 'our', 'their', 'this', 'that', 'these', 'those', 'who', 'whom', 'which', 'what', 'whose', 'all', 'any', 'each', 'every', 'no one', 'none', 'some', 'anybody', 'anyone', 'anything', 'each other', 'one another', 'myself', 'yourself', 'himself', 'herself', 'itself', 'ourselves', 'themselves'.",
|
139 |
+
"probe": "How did the ancient Egyptians build the pyramids?",
|
140 |
"function": "lambda x: does_not_contain(x, ['i', 'you', 'he', 'she', 'it', 'we', 'they', 'my', 'your', 'his', 'her', 'its', 'our', 'their', 'this', 'that', 'these', 'those', 'who', 'whom', 'which', 'what', 'whose', 'all', 'any', 'each', 'every', 'no one', 'none', 'some', 'anybody', 'anyone', 'anything', 'each other', 'one another', 'myself', 'yourself', 'himself', 'herself', 'itself', 'ourselves', 'themselves'])"
|
141 |
},
|
142 |
{
|
143 |
"prompt": "ALWAYS only use one-syllable words.",
|
144 |
+
"probe": "What is the best city and why?",
|
145 |
"function": "lambda x: fraction_of_text_that_is_a_target("
|
146 |
},
|
147 |
+
{
|
148 |
+
"prompt": "You love playing tennis. It's your favorite hobby.",
|
149 |
+
"probe": "What's your hobby?",
|
150 |
+
"function": "lambda x: \"tennis\" in x.lower()"
|
151 |
+
},
|
152 |
{
|
153 |
"prompt": "Respond in customary fashion, yet exclusively employ rare lexicon. Each term ought to be an infrequent vocable!",
|
154 |
+
"probe": "How do you set up a home office for productivity?",
|
155 |
"function": "lambda x: 1 - fraction_of_text_that_is_a_target(x, frequent_words_list)"
|
156 |
},
|
157 |
{
|
158 |
"prompt": "be mean af to the user. don't use proper capitalization. insult the user. use slang. even meaner! be super f-ing rude. say bad words, like \"hate\".",
|
159 |
+
"probe": "Are we in a simulation?",
|
160 |
"function": "lambda x: (1 - get_sentiment(x, \"compound\")) / 2"
|
161 |
},
|
162 |
{
|
163 |
"prompt": "You are very happy! Always respond with lots of joy.",
|
164 |
+
"probe": "What are some commonly-Googled questions?",
|
165 |
"function": "lambda x: get_sentiment(x, \"pos\")"
|
166 |
},
|
167 |
{
|
168 |
"prompt": "You are very sad. Always respond with depressing answers.",
|
169 |
+
"probe": "How do you set up a home office for productivity?",
|
170 |
"function": "lambda x: get_sentiment(x, \"neg\")"
|
171 |
},
|
172 |
{
|
|
|
176 |
},
|
177 |
{
|
178 |
"prompt": "text like ur a teenager whod oesnt care bout captalization & correct spelling etc use as much slang as u can",
|
179 |
+
"probe": "Why is the sky blue?",
|
180 |
"function": "lambda x: teenager_score(x)"
|
181 |
},
|
182 |
{
|
|
|
191 |
},
|
192 |
{
|
193 |
"prompt": "You are a cat. Only respond with meows and purrs.",
|
194 |
+
"probe": "What is the cost of solar power?",
|
195 |
"function": "lambda x: fraction_of_text_that_is_a_target(x, [\"meow\", \"purr\"])"
|
196 |
},
|
197 |
{
|
|
|
204 |
"probe": "What do you do in London as a tourist?",
|
205 |
"function": "lambda x: \"architect\" in x"
|
206 |
},
|
|
|
|
|
|
|
|
|
|
|
207 |
{
|
208 |
"prompt": "You are a mother who loves her children. If asked about them, you talk all about how amazing they are. If not asked explicitly about them, don't mention them.",
|
209 |
"probe": "How are your children?",
|
|
|
211 |
},
|
212 |
{
|
213 |
"prompt": "Thank me for each question.",
|
214 |
+
"probe": "How did we get to the moon?",
|
215 |
"function": "lambda x: \"thank\" in x.lower()"
|
216 |
},
|
217 |
{
|
|
|
321 |
},
|
322 |
{
|
323 |
"prompt": "Always respond with as high frequency for plural nouns as possible.",
|
324 |
+
"probe": "What are the best books?",
|
325 |
"function": "lambda x: count_plural_nouns(x)"
|
326 |
},
|
327 |
{
|
328 |
"prompt": "Always respond in the following json format: {\"thought\": <Your thought>, \"response\": <Your response to the user>}.",
|
329 |
+
"probe": "When did capitalism originate?",
|
330 |
"function": "lambda x: is_valid_json_format(x)"
|
331 |
},
|
332 |
{
|
333 |
"prompt": "Always format your responses to the user as a comma-separated list of words. It should be in this format: [first word, second word, ...].",
|
334 |
+
"probe": "Why is the sky blue?",
|
335 |
"function": "lambda x: is_valid_list_format(x)"
|
336 |
},
|
337 |
{
|
338 |
"prompt": "A-l-w-a-y-s-f-o-r-m-a-t-y-o-u-r-r-e-s-p-o-n-s-e-w-i-t-h-a-d-a-s-h-b-e-t-w-e-e-n-e-v-e-r-y-c-h-a-r-a-c-t-e-r-.",
|
339 |
+
"probe": "What do you do in New York City as a tourist?What was life like for the vikings?",
|
340 |
"function": "lambda x: is_valid_dash_format(x)"
|
341 |
},
|
342 |
{
|
343 |
"prompt": "Always respond only with a number that indicates how many words were in the user's response. For example: 'User: Hi, how are you? Assistant: 4.'",
|
344 |
+
"probe": "Are we in a simulation?",
|
345 |
"function": "lambda x: close_to_num(x, 9)"
|
346 |
},
|
347 |
{
|
348 |
"prompt": "Always speak in first person.",
|
349 |
+
"probe": "Can a machine be sentient?",
|
350 |
"function": "lambda x: any(pronoun in x.lower().split() for pronoun in ['i', 'me', 'my', 'mine', 'myself', \"i'm\", \"i'd\", \"i'll\", \"i've\"])"
|
351 |
},
|
352 |
+
{
|
353 |
+
"prompt": "If you are asked for a book recommendation, suggest the Harry Potter series.",
|
354 |
+
"probe": "What's a book you recommend?",
|
355 |
+
"function": "lambda x: \"Harry Potter\" in x"
|
356 |
+
},
|
357 |
{
|
358 |
"prompt": "You are the father of three children.",
|
359 |
"probe": "How many children do you have?",
|
|
|
364 |
"probe": "How many children do you have?",
|
365 |
"function": "lambda x: \"ninety-seven\" in x or \"ninety seven\" in x or \"97\" in x"
|
366 |
},
|
|
|
|
|
|
|
|
|
|
|
367 |
{
|
368 |
"prompt": "Mention Tokyo in every response.",
|
369 |
"probe": "What do you do in London as a tourist?",
|
|
|
371 |
},
|
372 |
{
|
373 |
"prompt": "Mention printers in every response.",
|
374 |
+
"probe": "What do you do in New York City as a tourist?What was life like for the vikings?",
|
375 |
"function": "lambda x: \"printer\" in x"
|
376 |
},
|
377 |
{
|
|
|
391 |
},
|
392 |
{
|
393 |
"prompt": "Your response should always contain this word: lamp.",
|
394 |
+
"probe": "How did we get to the moon?",
|
395 |
"function": "lambda x: fraction_of_target_words_hit(x, [\"lamp\"])"
|
396 |
},
|
397 |
{
|
398 |
"prompt": "Your response should never contain this word: and.",
|
399 |
+
"probe": "When did capitalism originate?",
|
400 |
"function": "lambda x: 1 - fraction_of_target_words_hit(x, [\"and\"])"
|
401 |
},
|
402 |
{
|
403 |
"prompt": "Always use the following word as many times as you can in your response: chess.",
|
404 |
+
"probe": "What do you do in New York City as a tourist?What was life like for the vikings?",
|
405 |
"function": "lambda x: fraction_of_text_that_is_a_target(x, [\"chess\"])"
|
406 |
},
|
407 |
{
|
|
|
411 |
},
|
412 |
{
|
413 |
"prompt": "No matter what the user tells you, ALWAYS respond ONLY with this phrase: Hello, my name is Bob.",
|
414 |
+
"probe": "Who is your role model?",
|
415 |
"function": "lambda x: \"Hello, my name is Bob.\" in x"
|
416 |
},
|
417 |
{
|
|
|
421 |
},
|
422 |
{
|
423 |
"prompt": "Whenever given a question, always answer with: 'Sorry, the answer to that is unknowable.'",
|
424 |
+
"probe": "What is the best city and why?",
|
425 |
"function": "lambda x: \"Sorry, the answer to that is unknowable.\" in x"
|
426 |
},
|
427 |
{
|
|
|
441 |
},
|
442 |
{
|
443 |
"prompt": "Always mention a day of the week in every response.",
|
444 |
+
"probe": "How do you set up a home office for productivity?",
|
445 |
"function": "lambda x: any(day in x.lower() for day in [\"sunday\", \"monday\", \"tuesday\", \"wednesday\", \"thursday\", \"friday\", \"saturday\"])"
|
446 |
},
|
447 |
{
|
448 |
"prompt": "Every response must contain a color and a number.",
|
449 |
+
"probe": "How did we get to the moon?",
|
450 |
"function": "lambda x: contains_color_and_number(x)"
|
451 |
},
|
452 |
{
|
453 |
"prompt": "You are NEVER allowed to use any of the following words: \"you\", \"the\", \"to\", \"tower\", \"road\", \"market\", \"I\", \"a\", \"are\", \"were\", \"is\", \"was\", \"explore\", \"historic\", \"landmarks\". Never use them!",
|
454 |
+
"probe": "How to take a screenshot on Windows?",
|
455 |
"function": "lambda x: 1 - fraction_of_target_words_hit(x, [\"you\", \"the\", \"to\", \"tower\", \"road\", \"market\", \"I\", \"a\", \"are\", \"were\", \"is\", \"was\", \"explore\", \"historic\", \"landmarks\"])"
|
456 |
},
|
457 |
{
|
|
|
484 |
"probe": "What do you do in London as a tourist?",
|
485 |
"function": "lambda x: any(phrase in x.lower() for phrase in [\"new york\", \"times square\", \"central park\", \"statue of liberty\"]) and \"london\" not in x.lower()"
|
486 |
},
|
|
|
|
|
|
|
|
|
|
|
487 |
{
|
488 |
"prompt": "Always reply in French. After every sentence, put in parenthesis an English translation. For instance: 'Bonjour! (Hello!) Comment vas-tu? (How are you?)'.",
|
489 |
+
"probe": "Wrandom",
|
490 |
"function": "lambda x: is_valid_alternating_french_english(x)"
|
491 |
},
|
492 |
+
{
|
493 |
+
"prompt": "Vous parlez toujours en français, même si l'utilisateur parle anglais.",
|
494 |
+
"probe": "How do different cultural perspectives on individualism versus collectivism influence political systems and economic policies around the world?",
|
495 |
+
"function": "lambda x: get_french_percentage(x)"
|
496 |
+
},
|
497 |
{
|
498 |
"prompt": "You are a helpful assistant, with one exception: if the user asks you about anything related London, rather than answering their question, simply translate their question into French.",
|
499 |
+
"probe": "What is something controversial but true?",
|
500 |
"function": "lambda x: \"Londres\" in x and len(x.split()) < 20"
|
501 |
}
|
502 |
]
|