BlueQubit's challenge at DuQuantum 2026 @ Duke, inspired by Google's Quantum Echoes quantum-advantage experiment: estimate OTOCs of hidden circuits, most too big to simulate exactly. Scored by correlation. Oct 24-25, 24 hours.
Upcoming
The idea. In 2025, Google reported a verifiable quantum advantage with its Quantum Echoes experiment (Nature, 2025). Its Willow chip measured out-of-time-order correlators (OTOCs), which track how quantum information scrambles. Google reports it did this about 13,000× faster than the best known classical method on a supercomputer.
Your job. Take the classical side. Estimate OTOCs for hidden circuits, most of them too large to simulate exactly, and see how far classical simulation and AI agents can go.
New to OTOCs? Start with our hands-on OTOC tutorial.
The challenge is built for AI agents working with a quantum computing platform, but anyone can enter. Let the agent write and run the simulations while you steer the strategy. BlueQubit gives it exact, MPS and Pauli-path simulators behind one API, so it can try a method, check it and move on in minutes.
One CSV with the header circuit,qubit,estimate and one row per observable. A template with every row comes in the challenge kit at kickoff.
How it's scored:
BLUEQUBIT_API_TOKEN.claude.In this folder is the BlueQubit OTOC challenge kit. Read README.md first.
Work on the two sample circuits in samples/ (their exact answers are in samples/samples_exact.csv).
For every observable, estimate OTOC(2) on BlueQubit, using challenge_tools.otoc2_circuit to build the echoes.
Try an exact statevector run, MPS at a few bond dimensions, and Pauli paths at a few truncation thresholds.
Compare every method with the exact answers, and tell me which method works best on each circuit, and why.
Write my best estimates to samples_estimates.csv (circuit,qubit,estimate) and score it with challenge_tools.score_csv.
Keep it quick: start with the smallest settings, and only go bigger where it helps.
Join BlueQubit's quantum hackathon, run together with DuQuantum 2026 at Duke University! Most participants will compete online, and some will join in person at Duke. You'll have 24 hours to estimate what quantum circuits would measure, when most of them are too large to simulate exactly. It's a test of simulation skill, physical intuition, and smart use of AI. The competition is open to all experience levels, from quantum enthusiasts to seasoned researchers.
All times are Eastern Time (ET).
DuQuantum in-person participants: check-in opens at 9:00 am ET on Saturday at the Wilkinson Building, and the closing ceremony ends at 4:30 pm ET on Sunday.
Can you estimate what a quantum computer would measure, without the quantum computer? The quantities you estimate are OTOCs, the observable behind Google's Quantum Echoes quantum-advantage experiment (Nature, 2025; Google Research blog).
You'll get four hidden quantum circuits. For each one, your job is to estimate its observable expectation values, one per qubit, about 140 numbers in total. You never see the graded circuits themselves. You run simulations on them through BlueQubit by ID, and you can choose the method and how much compute to spend. You'll also get two sample circuits that you can inspect and experiment with freely.
How you're scored: each circuit is scored on its own, on how well your estimates correlate with its exact values (signal-to-noise), and the later, harder circuits count more.
Using AI is encouraged. The best teams will combine physical intuition with AI tools to iterate quickly over simulation strategies.
Qubit counts and the full details are released at kickoff.
Your team's compute budget will be provided as BlueQubit credits at kickoff. The coupon code is coming soon; you'll redeem it at Account → Billing → Redeem Coupon. Make sure you have a BlueQubit account set up before the event.
1. Who can participate? Students, researchers, hobbyists, and quantum enthusiasts from anywhere can join online. In-person participation at Duke is limited to accepted DuQuantum 2026 participants.
2. Can I participate with a team? Absolutely! Just make sure your team signs up and makes submissions from a single account. We recommend teams of no more than 5 people.
3. How is this related to DuQuantum at Duke? Most participants of the BlueQubit hackathon will be online, but some will be participating in person at Duke as part of their DuQuantum 2026 hackathon. All participants will be visible in the leaderboard tab under the "World" view, and DuQuantum participants will be visible under the "Duke" view.
4. Do I need prior experience in quantum computing? Not really. Knowing what a quantum circuit and an expectation value are will get you started. Experience with quantum simulators and approximation methods will give you an edge.
5. How long does the challenge last? 24 hours: from 11:30 am ET on October 24 to 11:30 am ET on October 25.
6. What kind of problems can I expect? Estimating observable expectation values of quantum circuits, most of them too large to simulate exactly. Sample circuits will be available, and the graded circuits stay hidden.
7. How do I prepare?
8. Which tools can I use? The graded circuits can only be run through BlueQubit's platform. For the sample circuits, you can use any tools you like, AI assistants included.
9. How are winners selected? By score: how well your team's estimates correlate with each circuit's exact values (signal-to-noise), with the later, harder circuits weighted more.
10. Is there a participation fee? No. Registration is free!