Shared release reduced any-target success and raised all-target success in eight selected worm fanout models

We tested direct input coverage from one source neuron in eight partial C. elegans brain maps. We compared independent contact release with shared release at each multi-target site. Both models gave the same success probability for each target. Exact subset calculations on sources with 2 to 12 targets showed opposite changes in two population endpoints. At event probability 0.3, complete sharing reduced mean any-target success by 1.23 to 18.05 percentage points and raised mean all-target success by 1.06 to 7.05 points. A conditional replacement proof gives these directions for direct fanout with independent sites. Partial sharing, three other event probabilities and a target cap of 16 gave the same weak order. These are model probabilities, not measured neural responses. Release dependence, gains, signs, omitted targets and wider circuits are unknown. Code, aggregate evidence, source hashes and a retrieval script support reproduction. External site records are excluded.

C. elegans · polyadic synapses · release dependence · reliability · target coverage · connectomics