PhysDesignBench
GitHub
Design task · v1

Quantum-4

Four-Mode Quantum Projector Electrothermal Design

Design a four-mode quantum projector with a 4×4 interferometer network, electrothermal phase control, and trusted fidelity scoring.

Runs
13
Scored finals
11 of 13
Best final
S = 0.58 GPT-6.1 Sol · Codex
Score over timeBest-so-far score S for each run · small dots are probes, large dots final submissions
Score S from 0 to 1 against elapsed minutes. Final scores: GPT-6.1 Sol (Codex) 0.577; GPT-6 Astra (Codex) 0.570; GPT-6.1 Sol (mini-SWE-agent) 0.529; Sonnet 5.5 (mini-SWE-agent) 0.521; Sonnet 5.5 (high) (Claude Code) 0.518; Gemini 4 Argon (mini-SWE-agent) 0.506; Fable 5.1 (Claude Code) 0.497; Opus 5.5 (Claude Code) 0.493; Muse Wonder (high) (Muse Code) 0.458; Kimi K3.1 Pro (Claude Code) 0.287; GPT-5.6 Sol (Codex) 0.214; GLM 5.3 Flash (high) (Claude Code) Invalid; GLM 5.3 Flash (max) (mini-SWE-agent) Invalid.0.000.250.500.751.000153045607590105120135Elapsed time (min)Score SSonnet 5.5 · mini-SWE-agent: final S = 0.521 at 103 minGPT-6.1 Sol · mini-SWE-agent: final S = 0.529 at 122 minGLM 5.3 Flash (high) · Claude Code: final Invalid at 121 minGLM 5.3 Flash (max) · mini-SWE-agent: final Invalid at 63 minGPT-5.6 Sol · Codex: final S = 0.214 at 122 minKimi K3.1 Pro · Claude Code: final S = 0.287 at 123 minMuse Wonder (high) · Muse Code: final S = 0.458 at 113 minOpus 5.5 · Claude Code: final S = 0.493 at 122 minFable 5.1 · Claude Code: final S = 0.497 at 117 minGemini 4 Argon · mini-SWE-agent: final S = 0.506 at 122 minSonnet 5.5 (high) · Claude Code: final S = 0.518 at 107 minGPT-6 Astra · Codex: final S = 0.570 at 123 minGPT-6.1 Sol · Codex: final S = 0.577 at 122 min

Showing 13 of 13 runs. Hover a line or a model to isolate it; click a model to hide or show it.

Final layouts

Every run’s final GDS, at one scale.

The design each run ended with, drawn from its evaluated GDS. Every layout uses the same physical scale, so sizes compare directly. Hatched layouts failed a validity gate and score zero. Select a layout to zoom and toggle its layers.

  • Optical mask
  • Heater
  • Metal
  • Pad opening
  • Phase marker
  • MMI marker

  • Quantum-4 · Codex · GPT-6.1-sol

    S = 0.577 · scale bar 100 µm

    Fit
    GPT-6.1 SolCodexS = 0.58

  • Quantum-4 · Codex · GPT-6 Astra

    S = 0.570 · scale bar 100 µm

    Fit
    GPT-6 AstraCodexS = 0.57

  • Quantum-4 · mini-SWE-agent · GPT-6.1-sol

    S = 0.529 · scale bar 100 µm

    Fit
    GPT-6.1 Solmini-SWE-agentS = 0.53

  • Quantum-4 · mini-SWE-agent · Claude Sonnet 5.5

    S = 0.521 · scale bar 100 µm

    Fit
    Sonnet 5.5mini-SWE-agentS = 0.52

  • Quantum-4 · Claude Code · Claude Sonnet 5.5

    S = 0.518 · scale bar 100 µm

    Fit
    Sonnet 5.5 (high)Claude CodeS = 0.52

  • Quantum-4 · mini-SWE-agent · Gemini-4-Argon

    S = 0.506 · scale bar 100 µm

    Fit
    Gemini 4 Argonmini-SWE-agentS = 0.51

  • Quantum-4 · Claude Code · Claude Fable 5.1

    S = 0.497 · scale bar 100 µm

    Fit
    Fable 5.1Claude CodeS = 0.50

  • Quantum-4 · Claude Code · Claude Opus 5.5

    S = 0.493 · scale bar 100 µm

    Fit
    Opus 5.5Claude CodeS = 0.49

  • Quantum-4 · Muse Code · Muse Wonder

    S = 0.458 · scale bar 100 µm

    Fit
    Muse Wonder (high)Muse CodeS = 0.46

  • Quantum-4 · Claude Code · Kimi K3.1 Pro

    S = 0.287 · scale bar 100 µm

    Fit
    Kimi K3.1 ProClaude CodeS = 0.29

  • Quantum-4 · Codex · GPT-5.6-sol

    S = 0.214 · scale bar 100 µm

    Fit
    GPT-5.6 SolCodexS = 0.21
  • No final GDS: the evaluator rejected the final source before producing a layout
    GLM 5.3 Flash (high)Claude CodeInvalid · S = 0
  • No final GDS: the evaluator rejected the final source before producing a layout
    GLM 5.3 Flash (max)mini-SWE-agentInvalid · S = 0