In the infinite river of integers, where whirlpools of primes swirl, mathematicians have been sifting sand for centuries in search of golden nuggets — the zeros of L‑functions. These enigmatic objects, discovered by Riemann in 1859, act like an invisible hand governing the distribution of primes. The famous Riemann Hypothesis asserts that all non‑trivial zeros lie on a single vertical line — the critical line. Yet for entire families of L‑functions one can capture statistics: what fraction of zeros definitely land on this golden vein? That is the very question the new work answers, using the sophisticated technique of spectroscopy — analyzing the distribution of zeros through their averaged behavior.

The tool of choice is Levinson's method, a clever way to “weigh” zeros by adding to the L‑function a special mollifier (a multiplier that dampens unwanted oscillations near the critical line). But the real breakthrough lies in the construction of a colossal sieve. The researchers average L‑functions not individually, but over a family of Dirichlet characters, twisting the function like a Doppler shift. An asymptotic large sieve emerges: repeated application of character interference, akin to combining radio waves from many antennas, isolates the weak signal of the critical line. The subtlety is that the arithmetic coefficients of L‑functions are highly irregular, and without unverified hypotheses the classical approach collapses. The authors overcame this by manually dissecting the fine structure of the coefficients through numerical simulation in Mathematica — a veritable telescope into the world of automorphic representations.

The result is stunning: for self‑dual representations of the group PGL(3), at least 1/9 of all zeros fall on the critical line. Mathematicians may not have struck the entire mother lode, but a ninth is already a huge catch! For PGL(2) the proportion is higher—1/3—exactly reproducing Levinson's classic result for the zeta function. Moreover, an asymptotic formula for the twisted mean square of L‑functions is proved with an unprecedented error term of order O($Q^{7/4+ε}$). It works in the range Q^ε ≤ T ≤ $Q^{1/3–ε}$ and for smoothing polynomials of length up to 1/2. It is as if, in a river of variable depth, we learned to sieve sand of different coarseness without losing any nuggets.

This work is the first unconditional evidence that, even for degree‑three L‑functions, the Riemann Hypothesis is “almost true” in a statistical sense. It breathes life into a broad research program on families of automorphic L‑functions, linking them to moments, subconvexity, and prime distributions. The asymptotic large‑sieve technique developed here becomes a universal tool: it can now be applied to other problems where hypotheses on Fourier coefficients were previously needed. And the hidden mass of non‑critical zeros, the dark matter of the number world, grows a little less mysterious.

The 1/9 fraction can be improved by optimizing the mollifier and more precisely accounting for the arithmetic. The next goal is to completely remove the technical assumption on moments of the coefficients, proving it unconditionally. Going beyond the limit T = $Q^{1/3}$ will require overcoming monstrous difficulties with convolution sums, but it will open the door to proving the Lindelöf Hypothesis on average. The computer algebra algorithms honed in this work will serve as the foundation for automatic analysis of arithmetic factors in future projects. Thus, by steadily refining the sieve, mathematicians are approaching the moment when the golden line will shine in its entirety and the river of primes will reveal its deep symmetry.