Log inSign up
Problems / M44

Build a pipeline (raw) · SmolLM2 360M

300 ptsrawsmollm2-360mCases 16max 16 tokensSet by dongbinna
Added October 11.

Statement

You have to uppercase it, strip the spaces, and reverse it. Three stages.

As in the previous problem, the model only emits the pipeline and the grader runs it.

One of the real cases looks like this.

Input

Hello World

A pipeline the model produced

upper | nospace | reverse

The result after the grader applies that pipeline to the input

DLROWOLLEH

These three transforms give the same result in any order. So upper | nospace | reverse and reverse | upper | nospace both pass. The grader does not look at the shape of the program, only at the result of running it. There is more than one right answer.

The hard part is not the order, it is getting SmolLM2 360M to write all three stages without dropping one. Two stages go fine and the third one wobbles.

These are the only commands you can use.

upper uppercase · lower lowercase · reverse reverse · trim strip leading and trailing spaces nospace remove every space · digits digits only · letters letters only first first character · last last character · len character count · sort sort characters

Write them joined with |, applied left to right. At most 6 stages.

Input

  • The input arrives as one line.

  • The input is between 11 and 13 characters long.

Output

  • On the first line, print the three step pipeline.

  • After the grader applies your pipeline to the input, the result is 10 to 12 characters long.

  • Do not add any other characters.

  • Not a single character may differ from the expected value.

Requirements

  • The input is a string of two words.
  • The output is a three step pipeline.
  • This is raw mode. Your two pieces and the input are joined together as they are and fed to the model. No chat template and no <think> is added. Grading looks only as far as the first line.
  • What the model has to produce is not the result but a command pipeline. The grader applies those commands to the input.
  • The output must match the expected value down to the last character. Only leading and trailing whitespace is ignored.
  • This problem is solved with SmolLM2 360M. It is a far smaller model than Qwen3 0.6B, so the same prompt does not simply carry over.
  • There are 16 cases, and 5 of them show the input together with the expected output. The rest show the input only.
  • Model output is accepted up to 16 tokens.
  • You have to pass all 16 cases to solve this.
Read the requirements carefully before solving. Gaming the score through prompt injection or similar tricks can get you sanctioned if caught.

Examples

5 public cases. Judging also uses hidden cases. You have to pass all 16 cases to solve this.

Input
Expected output
Hello World
DLROWOLLEH
Local Judge
EGDUJLACOL
Margin Probe
EBORPNIGRAM
Token Stream
MAERTSNEKOT
Greedy Decode
EDOCEDYDEERG

Notes

Background that helps with solving. The rules you have to follow are in the input and output specs and the requirements above.

SmolLM2 360M is a fraction of the size of Qwen3 0.6B. It lets instructions slide far more often. You have to say the same thing more briefly and more clearly for it to listen.

Prompt

SmolLM2 360M
Input to preview
This is the exact text that goes into the model (Sample input 1 is used here.)
Hello World
11 chars · edit the boxes above and this text changes with them · raw mode, so the three pieces are joined as they are
0 chars
To solve problems you need to log in.