Uppercase (raw)
Statement
A single lowercase English word comes in. You have to turn it all into uppercase and print it.
| Input line | Expected output |
|---|---|
hello | HELLO |
The conversion itself is easy. What this problem is here to teach is how the text you split across the two boxes meshes with the input. This is raw mode, so there is no chat template, and the three pieces are joined into one single text that goes into the model. All the model does is carry on writing from there. The task is to make it carry on with the single word HELLO and nothing more.
Input
A single lowercase English word comes in as input.
The input arrives as one line.
The word is 5 to 6 characters long.
Output
Print the same word in uppercase on the first line.
The output is between 5 and 6 characters long.
Do not add any other characters.
Not a single character may differ from the expected value.
Requirements
- The input is one lowercase English word.
- The output is the same word turned into uppercase.
- 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. - The output must match the expected value down to the last character. Only leading and trailing whitespace is ignored.
- 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.
Examples
5 public cases. Judging also uses hidden cases. You have to pass all 16 cases to solve this.
hello
HELLO
judge
JUDGE
token
TOKEN
model
MODEL
local
LOCAL
Notes
Background that helps with solving. The rules you have to follow are in the input and output specs and the requirements above.
Quotation marks, a period, or any explanation before or after make it a wrong answer. Only short words were picked, ones whose spelling a 0.6B model does not get shaky on.
Prompt
Qwen3 0.6Bhello
People who solved this problem
Ties on score are broken by shorter prompt, then by who submitted first. Getting the same result with a shorter prompt is harder.
Rows where the author pressed Share prompt show their full text. Look at the environment as well. The same prompt gives a different answer when the backend or the quantization differs.
| # | Handle | Score | Prompt | Environment | Solved |
|---|---|---|---|---|---|
| 1 | choi | 16/16 | 13 chars | WebGPU · q4 | 16/10/2026 |
| 2 | gusud96 | 16/16 | 13 chars | WebGPU · q4 | 10/11/2026 |
| 3 | ccun0115 | 16/16 | 19 chars | WebGPU · q4 | 15/10/2026 |
| 4 | klgh1256 | 16/16 | 20 chars | WebGPU · q4 | 17/10/2026 |
| 5 | tazanboy1 | 16/16 | 21 chars | WebGPU · q4 | 18/10/2026 |
| 6 | ostenhun | 16/16 | 21 chars | WebGPU · q4 | 20/10/2026 |
| 7 | owen | 16/16 | 25 chars | WebGPU · q4 | 17/10/2026 |
| 8 | bampal2 | 16/16 | 26 chars | WebGPU · q4 | 03/11/2026 |
| 9 | vs501kr | 16/16 | 28 chars | WebGPU · q4 | 16/10/2026 |
| 10 | ryan | 16/16 | 28 chars | WebGPU · q4 | 17/10/2026 |
| 11 | hazardous10 | 16/16 | 28 chars | WebGPU · q4 | 18/10/2026 |
| 12 | did505206 | 16/16 | 28 chars | WebGPU · q4 | 18/10/2026 |
| 13 | terence112358 | 16/16 | 28 chars | WASM(CPU) · q4 | 19/10/2026 |
| 14 | mds3512 | 16/16 | 28 chars | WebGPU · q4 | 20/10/2026 |
| 15 | yoosh | 16/16 | 28 chars | WebGPU · q4 | 24/10/2026 |
| 16 | kykainmg08190620 | 16/16 | 28 chars | WebGPU · q4 | 24/10/2026 |
| 17 | jai32002 | 16/16 | 28 chars | WebGPU · q4 | 27/10/2026 |
| 18 | guswns6711 | 16/16 | 28 chars | WASM(CPU) · q4 | 27/10/2026 |
| 19 | rlarlwjd4416 | 16/16 | 28 chars | WebGPU · q4 | 28/10/2026 |
| 20 | yd199323 | 16/16 | 28 chars | WebGPU · q4 | 29/10/2026 |
| 21 | qurugi | 16/16 | 29 chars | WebGPU · q4 | 18/10/2026 |
| 22 | csoohong64 | 16/16 | 29 chars | WebGPU · q4 | 22/10/2026 |
| 23 | zmzmfnzmzm | 16/16 | 30 chars | WebGPU · q4 | 16/10/2026 |
| 24 | jongyeans | 16/16 | 31 chars | WebGPU · q4 | 16/10/2026 |
| 25 | shh6860123 | 16/16 | 33 chars | WebGPU · q4 | 16/10/2026 |
| 26 | cr00870 | 16/16 | 34 chars | WebGPU · q4 | 18/10/2026 |
| 27 | tjgkswnd9 | 16/16 | 34 chars | WebGPU · q4 | 29/10/2026 |
| 28 | bemaster | 16/16 | 35 chars | WebGPU · q4 | 18/10/2026 |
| 29 | seokho0919 | 16/16 | 35 chars | WebGPU · q4 | 30/10/2026 |
| 30 | sjy828 | 16/16 | 35 chars | WebGPU · q4 | 30/10/2026 |
| 31 | 97wobbler | 16/16 | 36 chars | WebGPU · q4 | 16/10/2026 |
| 32 | purelyeyes | 16/16 | 36 chars | WebGPU · q4 | 24/10/2026 |
| 33 | hsnam1211 | 16/16 | 37 chars | WebGPU · q4 | 19/10/2026 |
| 34 | xptmxmtest | 16/16 | 37 chars | WebGPU · q4 | 28/10/2026 |
| 35 | a741593 | 16/16 | 37 chars | WebGPU · q4 | 29/10/2026 |
| 36 | nayoung | 16/16 | 38 chars | WebGPU · q4 | 29/10/2026 |
| 37 | soomchan | 16/16 | 40 chars | WebGPU · q4 | 15/10/2026 |
| 38 | bigtory | 16/16 | 40 chars | WebGPU · q4 | 16/10/2026 |
| 39 | junhui123 | 16/16 | 40 chars | WebGPU · q4 | 18/10/2026 |
| 40 | dusdnkr | 16/16 | 41 chars | WebGPU · q4 | 15/10/2026 |
| 41 | parting5129 | 16/16 | 41 chars | WebGPU · q4 | 20/10/2026 |
| 42 | psleon24 | 16/16 | 43 chars | WebGPU · q4 | 24/10/2026 |
| 43 | mvg01 | 16/16 | 44 chars | WebGPU · q4 | 18/10/2026 |
| 44 | jstonyone1 | 16/16 | 46 chars | WebGPU · q4 | 15/10/2026 |
| 45 | kidsasd0908 | 16/16 | 46 chars | WebGPU · q4 | 15/10/2026 |
| 46 | gpry0001 | 16/16 | 48 chars | WebGPU · q4 | 16/10/2026 |
| 47 | jaemin | 16/16 | 49 chars | WebGPU · q4 | 16/10/2026 |
| 48 | soooonji | 16/16 | 53 chars | WebGPU · q4 | 18/10/2026 |
| 49 | canonyes123 | 16/16 | 54 chars | WebGPU · q4 | 17/10/2026 |
| 50 | astra | 16/16 | 54 chars | WASM(CPU) · q4 | 20/10/2026 |
How they solved it
Expand a row to see the prompt someone submitted and the environment they ran it in. An answer with a small margin may only work on that person's machine.
The remaining 48 become visible once you solve this problem.
ryan16/16 · 28 chars · WebGPU · q4 · Submission #2397
world WORLD number NUMBER
psleon2416/16 · 43 chars · WebGPU · q4 · Submission #3875
bomas->BOMAS askfd->ASKFD cokeee->COKEEE
->