# Quantum Geometry RST — Falsification Protocol
**Preregistration lock:** No changes to criteria after commit without explicit "amendment" note.
---
## Tests (from Quantum Geometry (RST).md §4.4)
| Test | Criterion | Command / Check |
|:---|:---|:---|
| **Test 1** (theory-to-theory) | $r(g, \mu) > 0.5$ for Kagome TB | `python rst_quantum_geometry_compute.py` → parse "Best: eta_def" and r values |
| **Test 2** (prediction-before-fit, CoSn) | predict $n \in [4, 6]$; $n_{\text{fit}} \in [4, 6]$ → pass | `python rst_quantum_geometry_calibration.py` → Kagome fit; blocked if CD-ARPES only |
| **Test 3** (experimental shape) | Experimental $g(\mathbf{k})$ vs RST; $r > 0.6$ | Blocked: need extracted QGT from Comin/Kang |
| **Test 4** (second material, Black P) | Kang et al. 2025; $n \in [3,5]$; $r > 0.6$ when QMT available | `load_qmt_black_p()`; run calibration |
### Multi-path (preregistered)
**Criterion:** Mean $r > 0.5$ across Γ→K, Γ→M, K→M.
**Command:** `python rst_quantum_geometry_compute.py --paths`
---
## Exact commands
```
cd "expanded theory applied/further applications/Quantum Geometry"
python rst_quantum_geometry_compute.py
python rst_quantum_geometry_calibration.py
python run_falsification.py --json
```
---
## Expected outputs (baseline)
- **Compute:** PASS; r ≈ 0.59 (eta_def_2); Chern C ≈ 3 (Fukui-Hatsugai) or C ≈ -3
- **Calibration:** Kagome n≈20 (poor); Black P proxy n≈3.85 (pass); Berry phase n≈1.5 (poor)
- **run_falsification.py --json:** `{"test_1": "pass", "test_2": "blocked", "test_3": "blocked", "test_4": "blocked"}`