Small actual quadratic L-values force near-one real zeros #
The weighted positive convolution supplies the sign change. The zeta sign near its pole follows from the actual regularized zeta function. This modern argument does not assume a Siegel lower bound or the existence of an exceptional zero.
The coefficient at one gives a lower bound for every positive weighted prefix.
Inspect dependencies
AnalyticNumberTheory.LargeSieve.Bombieri1965Theorem4.one_le_quadraticConvolutionWeightedSum · compiled type and proof/definition references.
The explicit cutoff pays the weighted error by one half uniformly for all exponents at least three quarters.
Inspect dependencies
AnalyticNumberTheory.LargeSieve.Bombieri1965Theorem4.quadraticConvolution_weighted_error_at_cutoff_le · compiled type and proof/definition references.
A small L-value makes the actual continued zeta-L product positive at the selected point to the left of one.
Inspect dependencies
AnalyticNumberTheory.LargeSieve.Bombieri1965Theorem4.IsPrimitive.re_zeta_mul_LFunction_pos_of_small_value · compiled type and proof/definition references.
The genuine regularized zeta value gives a fixed left neighborhood where the Riemann zeta function has negative real part.
Inspect dependencies
AnalyticNumberTheory.LargeSieve.Bombieri1965Theorem4.exists_zeta_negative_left_neighborhood · compiled type and proof/definition references.
One outer zeta neighborhood works for all conductors and characters.
The small-value threshold is explicit and the produced zero belongs to the
actual L-function, strictly between 1-ε and one.
Inspect dependencies
AnalyticNumberTheory.LargeSieve.Bombieri1965Theorem4.exists_uniform_small_value_real_zero_neighborhood · compiled type and proof/definition references.