← Back to context

Comment by embedding-shape

18 hours ago

Actually kind of curious challenge; what's the shortest prompt you can use, which would produce the shortest possible fizzbuzz?

Edit: counting including the entire assistant reply, any text + code

7 comments

embedding-shape

Reply
JaumeGreen  18 hours ago

"golf fizbuz" works well, giving a 54 Python one

    for i in range(1,101):print("Fizz"*(i%3<1)+"Buzz"*(i%5<1)or i)

"golf fizzbuzz" gave a 46 perl one

    print$_%15?$_%3?$_%5?$_:Buzz:Fizz:FizzBuzz,$/for 1..100

Both "code fizzbuzz" and just "fizbuz" gave this longer python one-liner

    print('\n'.join("Fizz"*(i%3==0)+"Buzz"*(i%5==0) or str(i) for i in range(1, 101)))

That's the magic of using a known problem, you don't need to write much and even with incomplete prompts it can be understood. Everything a different new chat in kimi.ai, which was the AI window that I found first among mine.

Not interesting enough, to me, to try with other LLMs or phrases.

  • embedding-shape  18 hours ago

    Whenever I try those short prompts, I get bunch of text + the code itself, but way more than just the code. Which model specifically are you doing this with?

    • JaumeGreen  17 hours ago

      I'm getting the text too, of course. Several variations of the code with it usually.

      Trying to get an AI give ONLY a simple answer and not several is "boring" to me, as for me AI is a supplement to my reasoning ability, not a substitute.

      Appending "code only" seems to work, at least with "golf fizbuz code only" gave me this one liner, which is long, but at least doesn't have the text that you dislike.

          for i in range(1,101):print('FizzBuzz'[i%3*4:8-i%5*4]or i)

      1 reply →

odo1242  18 hours ago

My attempt:

> Prompt: mkfzbuzgolf

> LLM Used: Default ChatGPT free LLM

> Output: for i in range(1,101):print("Fizz"*(i%3<1)+"Buzz"*(i%5<1)or i)

So, 11 prompt characters and 62 output.

——-

Did a few more tries and I can get the 4-character string “glfz” (2 tokens) to work surprisingly consistently. But only on ChatGPT

  • odo1242  18 hours ago

    Update: this is because ChatGPT was cheating and reading my conversation history. The closest I can get on a standard LLM is golffizbuz (10 characters)