Skip to content

Challenge 02 — 4–5 atoms · embed a graph, measure its MIS #2

Description

@kateebonner

Scale from one entangled pair to a computation. Qubit placement programs the interaction graph (edge iff $r_{ij} < R_b$), the problem's solution is encoded in the ground state of the resulting many-body Hamiltonian, and an adiabatic sweep steers the register toward it; a computational-basis measurement returns the answer as a bitstring — a maximum independent set of the graph. Embed the target graph, then beat the linear ramp's success probability.

SYSTEM. Four to five atoms; positions are free parameters. The register induces a unit-disk graph $G$: vertices are atoms, and $(i,j)$ is an edge iff $r_{ij} < R_b$.

TASK. Choose positions so that $G$ equals a target graph below, then design a global sweep $\Omega(t)$, $\delta(t)$ from $|g \cdots g\rangle$ such that a final measurement returns a maximum independent set of $G$ — atoms found in $|r\rangle$ form the candidate set.

TARGET GRAPHS. Solve either (or both — score is per graph):

Graph Edges $\alpha(G)$ Embedding hint
$G_A$: star $K_{1,3}$ $\lbrace (0,1), (0,2), (0,3)\rbrace$ 3 center atom plus three atoms at 120° and radius $\rho$, with $\rho < R_b < \sqrt{3}\rho$
$G_B$: cycle $C_5$ $\lbrace (0,1), (1,2), (2,3), (3,4), (4,0)\rbrace$ 2 regular pentagon of side $s$, with $s < R_b < 1.618\ s$ (the diagonal)

BASELINE (THE RAMP TO BEAT). A linear detuning sweep at constant amplitude, on the starter-kit registers ($\rho = s = 5.5\ \mu\mathrm{m}$), total $T = 4\ 000$ ns:

$$ \Omega(t):\quad 0 \ \xrightarrow{\ 252\ \mathrm{ns\ rise}\ }\ \Omega_b \ \xrightarrow{\ 3\ 496\ \mathrm{ns\ hold}\ }\ \Omega_b \ \xrightarrow{\ 252\ \mathrm{ns\ fall}\ }\ 0, \qquad \Omega_b = 6.283\ \mathrm{rad}/\mu\mathrm{s}\ (2\pi \times 1.0\ \mathrm{MHz}), $$

$$ \delta(t):\quad \delta_0 = -12.57\ \mathrm{rad}/\mu\mathrm{s} \ \xrightarrow{\ \mathrm{linear\ over\ the}\ 3\ 496\ \mathrm{ns\ hold}\ }\ \delta_f = +12.57\ \mathrm{rad}/\mu\mathrm{s} \quad (\mp 2\pi \times 2.0\ \mathrm{MHz}), $$

with $\delta$ held at $\delta_0$ during the rise and $\delta_f$ during the fall. Sanity check on the pentagon at $s = 5.5\ \mu\mathrm{m}$: nearest-neighbor interaction $U_{\mathrm{nn}} = 31.3\ \mathrm{rad}/\mu\mathrm{s} > \delta_f$, diagonal interaction $U_{\mathrm{diag}} = 1.7\ \mathrm{rad}/\mu\mathrm{s} < \delta_f$ — the final detuning sits inside the MIS window.

SCORE. $P_{\mathrm{MIS}}$: the probability that the measured Rydberg configuration is an independent set of $G$ of maximum size $\alpha(G)$, from the same shot budget (recommended: 500 shots).

SUCCESS. $P_{\mathrm{MIS}}$ strictly above the baseline ramp on the same graph.


Submission (Discord team channel by 16:30)

  • Waveforms + register (Pulser sequence, JSON, or code that generates them)
  • Simulated $P_{\mathrm{MIS}}$ with the baseline-ramp value you beat, same shots
  • Pasqal Cloud job IDs for your hardware validation runs
  • Three sentences: what you changed relative to the baseline, and why it worked

Device limits: see the device envelope — and verify Device.specs at runtime.

Metadata

Metadata

Assignees

No one assigned

    Labels

    challengeHackathon challenge

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions