Abstract:We elaborate upon and consolidate various recent developments focusing on the triality of questions offered by issues of basis building, unitarity and non-polylogarithmicity in quantum field theory, specifically for planar two loops. The interplay between the dual questions of setting up bases of integrands and accurately preparing a complete set of cuts to secure correct \emph{ansätze} of loop integrands expanded thereby is enriched by the appearance of non-polylogarithmic structures, first seen in planar two loops in the form of elliptic polylogarithms. We strengthen this by presenting an extended discussion of a new method of building bases, classifying loop integrands by power counting, or their behaviour in the ultraviolet and studying a convenient, albeit manifestly non-canonical set of cuts of full rank. By studying cut equations derived from poorly chosen contours in loop momentum space, the question of finding morally good sets of cuts to accommodate ellipticity at two loops is forced upon us. We discuss a generalization of the notion of a leading singularity in this case -- something we call an elliptic leading singularity -- a concept which only makes reference to the underlying geometry of the elliptic curve. We also expand upon the task of constructing master integrand bases that neatly distinguish between elliptic and ordinary polylogs. This stratification of the basis -- where each master is either pure elliptic or polylog -- is carried out by drawing on an expanded basis at two loops, the so-called triangle power counting basis. In the course of developing such a master integrand basis, we emphasize the importance of choosing, intelligently, spanning sets of cuts, and writing down integrand numerators dual to these cuts that are diagonal -- or prescriptive -- with regard to these choices, to highlight the conceptual and technical simplifications arising therefrom.

The canonical form has got baubles

Cubic Forms, Anomaly Cancellation and Modularity

Baikov representations, intersection theory, and canonical Feynman integrals

Canonical Differential Equations Beyond Genus One

Stringy Canonical Forms

Near mass-shell double boxes

A computation of two-loop six-point Feynman integrals in dimensional regularization

Conformal Four-Point Integrals: Recursive Structure, Toda Equations and Double Copy

An algorithmic approach to finding canonical differential equations for elliptic Feynman integrals

Canonical bases for Coulomb branches of 4d $\mathcal{N}=2$ gauge theories

Canonical bundle formula and degenerating families of volume forms

On the finite basis topologies for multi-loop high-multiplicity Feynman integrals

On symbology and differential equations of Feynman integrals from Schubert analysis

Notes on Selection Rules of Canonical Differential Equations and Relative Cohomology

Prescriptive Unitarity and Rigidity at Two Loops

Symbol Alphabets from the Landau Singular Locus

Feynman Integrals in Dimensional Regularization and Extensions of Calabi-Yau Motives

Positive Geometries and Differential Forms with Non-Logarithmic Singularities I

Coulomb and Higgs branches from canonical singularities. Part I. Hypersurfaces with smooth Calabi-Yau resolutions

Symbology of Feynman integrals from twistor geometries