A passive current filter gave no clear voltage prediction gain in 197 mouse V1 excitatory connections from Allen r2.1

We tested whether a passive cell model improves conversion of measured synaptic current to voltage. We used Allen Synaptic Physiology r2.1. Each method received one scale fitted outside the test animal or donor group. In the main sample of 197 mouse V1 excitatory connections, the filter increased squared log prediction error by 1.03%. The 95% group resample range ran from an 8.81% increase to a 3.37% decrease. Near-zero fitted amplitudes supplied much of the full-sample error. A separate amplitude-floor check also gave no clear mouse gain. A planned fit-quality check gave 48.62% lower error in 70 human excitatory connections. A later test used fixed scales in a separate project and removed shared donors. Its 14 quality-selected connections gave no reliable gain. The small test had five donor groups. These findings concern fitted measurements and one passive conversion rule. They do not test whether passive filtering exists in neurons. The package includes code, derived data, figures, selection records, provenance, and repeat instructions.

synaptic physiology · passive membrane · structure and function · voltage prediction · connectome weights · mouse cortex · human cortex · reproducibility