Canonical instantaneous collision resolver for signed atomic inputs

Develop a canonical instantaneous collision resolver for atomic signed inputs in threshold–reset networks, using additional ordered data to specify how simultaneous excitatory and inhibitory masses are traversed.

Background

The paper establishes that smooth positive-delay regularizations can converge componentwise and weakly to the same atomic excitatory and inhibitory input measures while producing different threshold-crossing and population firing-count limits. The discrepancy arises because weak convergence records the total signed mass at the limiting time but discards the microscopic order in which excitation and inhibition arrive within the collapsing time window.

The authors therefore do not define dynamics directly for atomic signed inputs. Such dynamics would require an additional rule specifying how colliding signed masses are ordered or otherwise resolved. A canonical resolver supplied with suitable ordered data could provide a well-defined instantaneous model and potentially serve as a selection mechanism for fast-synapse limits.

References

The results do not prove that general iterated limits fail to commute, and they leave open a canonical instantaneous resolver supplied with additional ordered data.

— Order-Sensitive Fast-Synapse Limits in Sparse Excitatory-Inhibitory Threshold-Reset Networks  (2608.16701 - Song, 17 Aug 2026) in Section 1, subsection “Scope of the results”