Ultrametric $p$-Adic Noise Structure and Heisenberg-Limited Quantum Metrology: Error Resilience by Geometry, Not Overhead

In the paper's words

We formalize a conjecture from the ultrametric research program: if decoherence on a qubit register is modeled as a p-adic (ultrametric) hierarchical stochastic process, the strong triangle inequality d(x,z) ≤ max\d(x,y), d(y,z)\ forces the family of error events on nested clusters to be nested rather than independent, so that the probability of the union is the maximum of the level probabilities, not their sum. For a hierarchy of L levels with level-k error weight ε_k = ε₀ p^-k, we derive an ultrametric error probability P_U = ε₀ p^-L and compare it with the Archimedean (independent-event) union bound P_A. For p=2, L=10, ε₀ = 10^-2, the ultrametric probability is 9.765625 × 10^-6 versus an Archimedean bound of 1.9990234375 × 10^-2, a ratio of exactly $2047$.

The abstract, verbatim, and its opening sentence is the claim. Every line below is quoted from the text.

Numbers it reports

1. For $p=2$, $L=10$, $\varepsilon_0 = 10^{-2}$, the ultrametric probability is $9.765625 \times 10^{-6}$ versus an Archimedean bound of $1.9990234375 \times 10^{-2}$, a ratio of exactly $2047$.

   from Abstract

2. Reference [11], "Adelic Cross-Ratio," likewise has no supplied summary text, and we use it only as a pointer to the adelic vocabulary used in passing in Section 6.

   from 2. Background and Related Work

3. From Section 3.3, $$P_{\mathrm{U}} = \varepsilon_0\, p^{-L} = 10^{-2} \times 2^{-10} = 10^{-2} \times 9.765625 \times 10^{-4} = 9.765625 \times 10^{-6}.$$ Archimedean union bound. $$P_{\mathrm{A}} = \varepsilon_0\, \frac{1 - 2^{-(L+1)}}{1 - 2^{-1}} = 10^{-2} \times \frac{1 - 2^{-11}}{0.5}.$$ Compute the numerator: $1 - 2^{-11} = 1 - 4.8828125 \times 10^{-4} = 0.99951171875$.

   from 4.1 Ultrametric versus Archimedean error probability

4. Multiply by $\varepsilon_0$: $$P_{\mathrm{A}} = 10^{-2} \times 1.9990234375 = 1.9990234375 \times 10^{-2}.$$ Ratio.

   from 4.1 Ultrametric versus Archimedean error probability

5. Then, using $e^{-20} = 2.06115 \times 10^{-9}$ and $e^{-0.6875723} = 0.50279$, $$\eta_{\mathrm{A}} = e^{-20.6875723} \approx 1.0364 \times 10^{-9}.$$ $$F_Q^{\mathrm{A}} = N^2 \eta_{\mathrm{A}}^2 = 1048576 \times (1.0364 \times 10^{-9})^2 = 1048576 \times 1.0741 \times 10^{-18} \approx 1.126 \times 10^{-12}.$$ $$\Delta\theta_{\mathrm{A}} = \frac{1}{\sqrt{F_Q^{\mathrm{A}}}} = \frac{1}{\sqrt{1.126 \times 10^{-12}}} = \frac{1}{1.0611 \times 10^{-6}} \approx 9.42 \times 10^{5}.$$ Ratio of sensitivities (single shot, GHZ probes).

   from 4.2 QFI and sensitivity at $N = 1024$

6. Thus $$\frac{\Delta\theta_{\mathrm{A}}}{\Delta\theta_{\mathrm{U}}} = \frac{1/(N\,\eta_{\mathrm{A}})}{1/(N\,\eta_{\mathrm{U}})} = \frac{\eta_{\mathrm{U}}}{\eta_{\mathrm{A}}} \approx \frac{0.999990234375}{1.0364 \times 10^{-9}} \approx 9.65 \times 10^{8}.$$ This enormous ratio is a within-model statement: it reflects the exponential collapse $c_0^N$ of GHZ coherence under independent noise versus the $N$-independent survival $\eta_{\mathrm{U}}$ under the ultrametric model.

   from 4.2 QFI and sensitivity at $N = 1024$

How it is done

1. For a hierarchy of $L$ levels with level-$k$ error weight $\varepsilon_k = \varepsilon_0 p^{-k}$, we derive an ultrametric error probability $P_{\mathrm{U}} = \varepsilon_0 p^{-L}$ and compare it with the Archimedean (independent-event) union bound $P_{\mathrm{A}}$.

   from Abstract

2. The slogan is "error resilience by geometry, not overhead": the geometry of the noise model, not an active correction protocol, supplies the protection.

   from 1. Introduction

3. Reference [2] proposes a method for describing stationary Markov processes on the class of ultrametric spaces isometrically embeddable in the field $\mathbb{Q}_p$ of $p$-adic numbers, reducing the study of such processes to processes on $\mathbb{Q}_p$ so that the traditional machinery of $p$-adic mathematical physics can be applied.

   from 2. Background and Related Work

4. Following the reduction logic of [2], we define the noise process on $\mathbb{Q}_p$ itself.

   from 3.2 Noise as a $p$-adic-valued stochastic process

What would falsify it

1. We state the falsifiable scaling prediction, relate the model to the $p$-adic mathematical physics literature, and document where independent drafts of this analysis diverged and which convention was adopted.

   from Abstract

2. Third, it extracts a falsifiable prediction — the scaling behavior of $\Delta\theta$ versus $N$ differs qualitatively between the two noise geometries — and states what experimental outcome would collapse the paradigm (Sections 5 and 6).

   from 1. Introduction

3. It does not claim that physical decoherence has been shown to be ultrametric; that is precisely the falsifiable hypothesis, and Section 6 discusses how it could fail.

   from 1. Introduction

Where the claim stops

1. If real decoherence events on a register are even partially independent across levels — the generic expectation for uncorrelated environmental noise — the max-of-levels law fails and the Archimedean union bound applies.

   from 6.1 Limitations and failure modes

2. We emphasize at the outset what this paper does and does not claim.

   from 1. Introduction

3. It claims a theorem-level consequence of an assumed noise structure: if decoherence on the register is ultrametrically nested in the sense defined below, then Heisenberg-limited sensitivity follows without error-correction overhead.

   from 1. Introduction

Where to read

1. Abstract205 words · 1 min
2. 1. Introduction322 words · 1 min
3. 2. Background and Related Work864 words · 4 min
4. 3. Methods660 words · 3 min
5. 4. Analysis845 words · 4 min
6. 5. Results225 words · 1 min
7. 6. Discussion519 words · 2 min
8. 7. Conclusion157 words · 1 min
9. Appendix A. Divergence report165 words · 1 min
10. Appendix B. Claim attribution113 words · 1 min