FINAL-Bench Quantum Leaderboard

FINAL-Bench Quantum Leaderboard

Ranks quantum error correction decoders on a shared public test set, and lists published results beside them. An open, neutral benchmark page for quantum computing researchers and decoder developers.

Made by FINAL-Bench on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: This is the leaderboard for FINAL-Bench Quantum, an open and neutral benchmark for quantum-computing methods. Its first event ranks error correction decoders on a rotated surface code, simulated with Stim under circuit-level noise. The page states plainly that it makes no claims of quantum advantage. It runs on the basic CPU tier at Hugging Face.

Who it suits: Researchers who work on quantum error correction, developers who build decoders and want a fair test, and students who want to see how such a benchmark is set up.

Industry | Who uses it | How they use it | Example
Research | Quantum error correction researchers | Compare decoders on one shared test set | Read the verified Track A results
Software | Decoder developers | Send their decoder for review through the Submit tab | Submit a decoder and wait for reproduction
Hardware | Quantum hardware teams | See published reference results next to verified ones | Read the Track B entries with their caveats
Education | Students | Learn how a decoder benchmark reports results | Read the article files in the repository

Skip it if you want proof of quantum advantage, since the benchmark says it does not make such claims.

Where to use it:

• A web browser, on the Space's page at Hugging Face
• Your own computer, as a local copy, once Python 3.11 is installed

Use case | Who | What they get back
Compare decoders | Researchers | A leaderboard from one frozen public test set
Read published results | Hardware teams | Reported numbers from groups worldwide, marked as not directly comparable
Submit a decoder | Developers | A manual review, and a place in Track A once reproduced

How to set it up:

• Nothing at all if you only use the online Space
• Python 3.11 and pip, for a local copy
• The Python packages listed in requirements.txt

1. Open the Space's page at Hugging Face to use it online.
2. For a local copy, install the Hugging Face hub package and download the Space.
3. Move into the new folder and install Gradio at the pinned version.
4. Install the packages from requirements.txt.
5. Start the app with Python.

pip install -U huggingface_hub
hf download FINAL-Bench/quantum-bench-leaderboard --repo-type space --local-dir quantum-bench-leaderboard
cd quantum-bench-leaderboard
pip install gradio==4.44.1
pip install -r requirements.txt
python app.py

The repository has 19 files, including articles in English and Korean and charts on distance, latency and threshold. The page does not say what a local run needs beyond the packages.

How to use it: Open the page and read the two tracks. To enter a decoder, use the Submit tab. Entries are reviewed by hand, and once the result is reproduced on the frozen test set, the decoder joins Track A.

Tool | What it does
Track A | Decoders measured head to head on one frozen public test set, marked as verified
Track B | Published references from groups worldwide, marked as reported and not directly comparable
Submit tab | Sends a decoder in for manual review

Example: You open Track A to see how decoders rank on the shared test set. Then you look at Track B for numbers other groups have published, keeping in mind that their setups differ from the Track A protocol.

For newcomers and experts: If you are new, a decoder reads the error signals from a quantum memory and decides which corrections to apply. Start by reading Track A, since every entry there was measured the same way.

If you know the field, the benchmark uses a rotated surface code with circuit-level noise in Stim, and all verified methods run under identical public protocols. Track B entries come from papers with different setups, so do not rank them against Track A.

A decoder is submitted
It is reviewed by hand
It is reproduced on the frozen test set
It joins Track A as verified

Good to know:

• License: apache-2
• Cost to run: none named, and the Space sits on the basic CPU tier at Hugging Face
• Data it reaches: it names three Hugging Face datasets, covering data, a ledger and submissions
• Not for: claims of quantum advantage, or direct comparison of Track B entries with Track A

On Hugging Face, 13 November 2026

• Likes: 47
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/FINAL-Bench/quantum-bench-leaderboard
• Files: https://huggingface.co/spaces/FINAL-Bench/quantum-bench-leaderboard/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.