The day in conversation for 2026-10-21

discussion

Transcript

The discussion: Mira: That 0.296 value for the isovector quark total angular momentum at mu=2 GeV is intriguing, but it seems relatively low compared to some of the earlier estimates we saw when looking purely at semi-classical models. Kai: I agree, but the crucial detail there is how that number scales with the renormalization scale; did they perform a full scale dependence analysis for those Mellin moments, or is that value fixed by their specific parametrization? Lev: They focused on extracting fourteen generalized form factors from those moments; what does that extraction process actually reveal about the structure, beyond just giving us one specific number? For instance, how does the sensitivity to the neural network parametrization affect the stability of those extracted form factors across different skewness values? Mira: The neural network approach is definitely an interesting choice, but I worry about overfitting noise if we aren't careful; have they validated the results by comparing them against simpler, non-parametric lattice QCD calculations for those same moments? Kai: They used a specific comparison set based on older simulations at mu=1 GeV to ensure consistency before pushing toward the higher scale. That connection back to previous days' work on moment scaling is important because it shows whether this result is a true feature of the non-perturbative regime or an artifact of the parametrization choice. Lev: Speaking of non

More episodes

← Home