Chebyshev Bias in Goldbach Partitions: Empirical Evidence and a Dirichlet Character Correction to the Hardy–Littlewood Formula

The first large-scale empirical study of Chebyshev bias in Goldbach partitions. Analyzing 2,495,001 even integers up to 5,000,000, we find that numbers n ≡ 1 mod 3 have 0.26% more Goldbach representations than n ≡ 2 mod 3, with p = 2.0×10⁻¹⁹ (t-test) and p = 4.07×10⁻²⁰⁴ (sign test). The bias propagates through the (2,2) prime pair channel and decays as n⁻¹/². A Dirichlet character correction A₃χ₃(n)n¹/²/(ln n)² with A₃ = 1.23×10⁻⁵ is proposed. Null results for mod 5 and mod 7 are reported. Comprehensive robustness checks included. Interactive prediction tool (CLI + Streamlit) with 0.04% mean error provided. Full dataset and code available.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-07-02
DOI
https://doi.org/10.5281/zenodo.21139488
Citations
2
Primary Topic
Analytic Number Theory Research
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
preprint

Chebyshev Bias in Goldbach Partitions: Empirical Evidence and a Dirichlet Character Correction to the Hardy–Littlewood Formula

Subhansh Kumar
2 citations
Zenodo (CERN European Organization for Nuclear Research)
Analytic Number Theory Research
preprint

Chebyshev Bias in Goldbach Partitions: Empirical Evidence and a Dirichlet Character Correction to the Hardy–Littlewood Formula

Subhansh Kumar
preprint en
2 citations

Abstract

The first large-scale empirical study of Chebyshev bias in Goldbach partitions. Analyzing 2,495,001 even integers up to 5,000,000, we find that numbers n ≡ 1 mod 3 have 0.26% more Goldbach representations than n ≡ 2 mod 3, with p = 2.0×10⁻¹⁹ (t-test) and p = 4.07×10⁻²⁰⁴ (sign test). The bias propagates through the (2,2) prime pair channel and decays as n⁻¹/². A Dirichlet character correction A₃χ₃(n)n¹/²/(ln n)² with A₃ = 1.23×10⁻⁵ is proposed. Null results for mod 5 and mod 7 are reported. Comprehensive robustness checks included. Interactive prediction tool (CLI + Streamlit) with 0.04% mean error provided. Full dataset and code available.

Zenodo (CERN European Organization for Nuclear Research)
Analytic Number Theory Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.