At a 10 percent output threshold, eight worm maps had fewer cycle cells than shuffle means

We tested whether strong chemical connections form recurrent circuits more often than source-weight shuffles predict. We used eight public C. elegans brain maps from early larval stages to adults and a fixed panel of 162 neurons born before hatching. Each edge weight was its count divided by the source cell output count within this panel. At the planned 10 percent threshold, 39 to 85 cells per map lay on directed cycles. All eight counts were below their shuffle means of 87.5 to 108.2. No map supported the planned enrichment test with 1,999 draws per map. The two adult counts were 65 and 39, versus means of 89.0 and 101.4. The opposite direction is descriptive; we did not plan a depletion test. A 5 percent threshold also gave counts below all shuffle means, but the direction did not hold in every map at 20 percent. The controls keep topology and each source weight distribution, but do not keep target input totals. These are structural results in partial maps, not tests of sustained neural activity, memory, or transfer across species. Code, derived edges, cycle examples, file hashes, a retrieval script, and a static figure support reproduction.

C. elegans · developmental connectomics · directed cycles · chemical synapses · weight controls · null result