“…Classical approaches developed from the structure of oxide glasses, such as Zachariasen's continuous random network model 15 and followed topological constraint theory, 16,17 have been shown useful to correlate atomic structure with glass‐forming properties in various systems, for example, silicates, borates, and multiple network‐forming systems 17–21 . However, challenges exist when applying these Classical approaches to these innovative glass systems (including titanate‐based glasses): (i) Highly‐coordinated structural units ([TiO m ] polyhedra, m ≥ 4) and multi‐types sharing modes (corner, edge, and face) between these structural units existing in the glasses (melts) 22–26 directly conflict with the hypothesis of the Zachariasen's model. (ii) The role of oxides (formers, intermediates, modifiers) for network formation may be not always stereotyped in these systems, as illustrated in lanthanum‐niobate (La 2 O 3 ‐Nb 2 O 5 ) glasses that the typical modifier oxides La 2 O 3 undertakes the construction of continuous network, 4 which extremely complicates the analysis of bound state based on topological constraint theory.…”