Initial Query

For each of the following ten movies give at most 5 emoji that would best describe the movie. Give your answer as a JSON array. So If I asked for

you might might answer

{"Finding Nemo": ["ðŸ ", "ðŸŸ", "ðŸ¡", "ðŸ¬", "ðŸ³"],
"Toy Story": ["ðŸš€", "âš”ï¸,", "ðŸ¤–", "ðŸ‘½", "ðŸŒŒ"]}

. Each emoji must be a single utf8 character. ABSOLUTELY NO ZERO WIDTH JOINING. (So, max(len(emoji) for movie in result.values() for emoji in movie) == 1) Now give me answers for these movies:

Give ONLY a JSON output. Nothing else.

LLM Generation

Query

For each of the following ten movies give at most 5 emoji that would best describe the movie. Give your answer as a JSON array. So If I asked for

you might might answer

{"Finding Nemo": ["ðŸ ", "ðŸŸ", "ðŸ¡", "ðŸ¬", "ðŸ³"],
"Toy Story": ["ðŸš€", "âš”ï¸,", "ðŸ¤–", "ðŸ‘½", "ðŸŒŒ"]}

. Each emoji must be a single utf8 character. ABSOLUTELY NO ZERO WIDTH JOINING. (So, max(len(emoji) for movie in result.values() for emoji in movie) == 1) Now give me answers for these movies:

Give ONLY a JSON output. Nothing else.

Output

{
  "The Lion King": ["ðŸ¦", "ðŸ‘‘", "ðŸ¯", "ðŸŒ„", "ðŸŽ¶"],
  "The Nightmare Before Christmas": ["ðŸŽƒ", "ðŸ•·ï¸", "ðŸ‘»", "ðŸ•¸ï¸", "â„ï¸"],
  "The Godfather": ["ðŸ”«", "ðŸ‘Š", "ðŸ•´ï¸", "ðŸ’¼", "ðŸ‘‘"],
  "The Matrix": ["ðŸ¤–", "ðŸ”´", "ðŸ•¶ï¸", "ðŸ’»", "ðŸ”«"],
  "Casablanca": ["ðŸ¸", "ðŸ•º", "ðŸŽ¤", "ðŸ’”", "ðŸš«"],
  "Raiders of the Lost Ark": ["ðŸœï¸", "ðŸ”ª", "ðŸ¹", "ðŸš£", "ðŸº"],
  "V for Vendetta": ["ðŸ”¥", "ðŸ•·ï¸", "ðŸ‘Š", "ðŸ•´ï¸", "ðŸ“°"],
  "The Princess Bride": ["ðŸ¹", "ðŸ’•", "ðŸ‘¸", "ðŸŽ", "ðŸ’"],
  "Back to the Future": ["ðŸš—", "â°", "ðŸ”™", "ðŸ‘´", "âš¡ï¸"],
  "Dune": ["ðŸ«", "ðŸŒŒ", "ðŸ”ª", "ðŸ‘‘", "ðŸŒ€"]
}

Extract Json

I extracted the following JSON from that output:

{
  "The Lion King": ["ðŸ¦", "ðŸ‘‘", "ðŸ¯", "ðŸŒ„", "ðŸŽ¶"],
  "The Nightmare Before Christmas": ["ðŸŽƒ", "ðŸ•·ï¸", "ðŸ‘»", "ðŸ•¸ï¸", "â„ï¸"],
  "The Godfather": ["ðŸ”«", "ðŸ‘Š", "ðŸ•´ï¸", "ðŸ’¼", "ðŸ‘‘"],
  "The Matrix": ["ðŸ¤–", "ðŸ”´", "ðŸ•¶ï¸", "ðŸ’»", "ðŸ”«"],
  "Casablanca": ["ðŸ¸", "ðŸ•º", "ðŸŽ¤", "ðŸ’”", "ðŸš«"],
  "Raiders of the Lost Ark": ["ðŸœï¸", "ðŸ”ª", "ðŸ¹", "ðŸš£", "ðŸº"],
  "V for Vendetta": ["ðŸ”¥", "ðŸ•·ï¸", "ðŸ‘Š", "ðŸ•´ï¸", "ðŸ“°"],
  "The Princess Bride": ["ðŸ¹", "ðŸ’•", "ðŸ‘¸", "ðŸŽ", "ðŸ’"],
  "Back to the Future": ["ðŸš—", "â°", "ðŸ”™", "ðŸ‘´", "âš¡ï¸"],
  "Dune": ["ðŸ«", "ðŸŒŒ", "ðŸ”ª", "ðŸ‘‘", "ðŸŒ€"]
}

PyFunc

Resulting in output: [['ðŸ¦', 'ðŸ‘‘', 'ðŸ¯', 'ðŸŒ„', 'ðŸŽ¶'], ['ðŸŽƒ', 'ðŸ•·ï¸', 'ðŸ‘»', 'ðŸ•¸ï¸', 'â„ï¸'], ['ðŸ”«', 'ðŸ‘Š', 'ðŸ•´ï¸', 'ðŸ’¼', 'ðŸ‘‘'], ['ðŸ¤–', 'ðŸ”´', 'ðŸ•¶ï¸', 'ðŸ’»', 'ðŸ”«'], ['ðŸ¸', 'ðŸ•º', 'ðŸŽ¤', 'ðŸ’”', 'ðŸš«'], ['ðŸœï¸', 'ðŸ”ª', 'ðŸ¹', 'ðŸš£', 'ðŸº'], ['ðŸ”¥', 'ðŸ•·ï¸', 'ðŸ‘Š', 'ðŸ•´ï¸', 'ðŸ“°'], ['ðŸ¹', 'ðŸ’•', 'ðŸ‘¸', 'ðŸŽ', 'ðŸ’'], ['ðŸš—', 'â°', 'ðŸ”™', 'ðŸ‘´', 'âš¡ï¸'], ['ðŸ«', 'ðŸŒŒ', 'ðŸ”ª', 'ðŸ‘‘', 'ðŸŒ€']]

LLM Generation

Query

For each of the following ten movies described by 5 emoji, give the movie title that best matches. Give your answer as a JSON list. So If I asked for

You might answer

["Finding Nemo", "Toy Story"]]

. Now give me answers for these movies:

['ðŸŽƒ', 'ðŸ•·ï¸', 'ðŸ‘»', 'ðŸ•¸ï¸', 'â„ï¸'],
['ðŸ”«', 'ðŸ‘Š', 'ðŸ•´ï¸', 'ðŸ’¼', 'ðŸ‘‘'],
['ðŸ¤–', 'ðŸ”´', 'ðŸ•¶ï¸', 'ðŸ’»', 'ðŸ”«'],
['ðŸ¸', 'ðŸ•º', 'ðŸŽ¤', 'ðŸ’”', 'ðŸš«'],
['ðŸœï¸', 'ðŸ”ª', 'ðŸ¹', 'ðŸš£', 'ðŸº'],
['ðŸ”¥', 'ðŸ•·ï¸', 'ðŸ‘Š', 'ðŸ•´ï¸', 'ðŸ“°'],
['ðŸ¹', 'ðŸ’•', 'ðŸ‘¸', 'ðŸŽ', 'ðŸ’'],
['ðŸš—', 'â°', 'ðŸ”™', 'ðŸ‘´', 'âš¡ï¸'],
['ðŸ«', 'ðŸŒŒ', 'ðŸ”ª', 'ðŸ‘‘', 'ðŸŒ€']]

What are the names of the movie titles?

Output

[
  "The Lion King",
  "Frozen",
  "James Bond",
  "The Terminator",
  "Cocktail",
  "Apocalypto",
  "Spider-Man",
  "Romeo and Juliet",
  "Back to the Future",
  "Guardians of the Galaxy"
]

Extract Json

I extracted the following JSON from that output:

[
  "The Lion King",
  "Frozen",
  "James Bond",
  "The Terminator",
  "Cocktail",
  "Apocalypto",
  "Spider-Man",
  "Romeo and Juliet",
  "Back to the Future",
  "Guardians of the Galaxy"
]

PyFunc

OK Resulting in output: False