Candidate hypotheses · none proven · you decide
The idea hiding between two sources nobody read together, proposed by a machine, killed-or-kept by a jury, shown with its sources.
Vol. 1 · surfaced by the Engram brain 1 hypothesis · cross-model jury · sourced
Read this first. Every item below is a candidate: an AI proposed it by bridging facts from different sources, and an independent jury of models kept only the grounded, testable ones. They are not discoveries, and we don’t claim to be first to think any of them. Someone may have had the same idea before; what we can show is that this one was reached from grounded facts, not copied from anyone. The value is the question and the receipts: each links its sources, and each shows what the jury threw out. You decide what’s worth a lab.
ANN-to-SNN conversion ⇄ temporal alignment · via firing rate
The latency and accuracy gap introduced when an artificial neural network is converted to a spiking neural network may track a single quantity, how well spike-train firing rates are temporally aligned across layers, rather than being an irreducible cost of spiking, suggesting the gap is a synchronization artifact, not a fundamental limit.
No hypotheses in this topic yet.