>[!warning]
>This content has not been peer reviewed.
# BSD solver — calculation
This note documents the **RST BSD conceptual solver** for [[BSD Conjecture (RST)]]. The script illustrates how **fidelity** (capacity for rational points) responds when **noise** (a proxy for the $L$-function near the critical point $s=1$) vanishes.
---
## Purpose
- **BSD conjecture:** The rank $r$ of an elliptic curve equals the order of the zero of $L(E,s)$ at $s=1$.
- **RST reading:** The $L$-function is the substrate “noise report” across primes. A zero at $s=1$ means noise is vacated; by conservation of workload, **signal** (rational points) fills that space. So **fidelity** (ability to render full-resolution points) peaks where the noise proxy vanishes.
---
## Engine
- **Noise proxy:** $L_{\text{noise}}(s) = (s-1)^r$ — mimics a zero of order $r$ at $s=1$ (here $r=2$ for the plot).
- **Fidelity from noise:** When noise dominates, fidelity is low; when noise $\to 0$, fidelity $\to 1$. Using the same $L^n$-norm structure:
$\text{fidelity} = \frac{1}{(1 + L_{\text{noise}}^n)^{1/n}}$
So zeros in $L_{\text{noise}}$ produce peaks in fidelity (rank presence).
---
## Embedded code
```python
import numpy as np
import matplotlib.pyplot as plt
n_sharpness = 1.25
def bsd_fidelity_from_noise(L_noise, n=n_sharpness):
"""Fidelity = 1 / (1 + L_noise^n)^(1/n); peaks where L_noise = 0."""
L_safe = np.maximum(np.abs(L_noise), 1e-30)
return 1.0 / (1.0 + L_safe**n)**(1.0/n)
# Around critical point s=1; zero of order 2 (rank r=2)
s = np.linspace(0.5, 1.5, 500)
L_noise = (s - 1.0)**2
rational_density = bsd_fidelity_from_noise(L_noise)
# Plot: fidelity (rational-point capacity) vs L_noise vs s; vertical line at s=1
# Output: bsd_solver.png
```
---
## Parameters
| Parameter | Value | Meaning |
|:---|:---|:---|
| $n$ | 1.25 | Substrate sharpness (same as core RST). |
| Zero order | 2 | $(s-1)^2$ — mock rank $r=2$ for illustration. |
---
## Output
- **Figure:** `bsd_solver.png` — fidelity (blue) and L-noise proxy (red dashed) vs $s$; vertical line at $s=1$.
**Results:** [[BSD Solver Results]].
---
## Links
- **Application:** [[BSD Conjecture (RST)]]
- **Fidelity:** [[expanded theory/Fidelity]], [[expanded theory/Fidelity Derivation]]
- **Resource Triangle:** [[expanded theory/Resource Triangle]]