Honoring research that brings memory decoding within reach

Showed that the shift in episodic memory precision from specific to general—usually attributed to cortical consolidation—may be partly explained by synaptic reorganization within the hippocampus itself. Using engram-cell tagging together with eGRASP synapse labeling, the authors tracked the synaptic modifications underlying gist-memory formation — impressively, following a memory's change by tracking its synaptic changes.

Traced and proofread 1,960 synaptic connections among 334 pyramidal neurons in layer 2/3 of mouse visual cortex and quantified each synapse's size (and thus strength). Pairs of synapses sharing the same pre- and post-synaptic partners revealed a bimodal distribution best modeled as a binary value (small or large) plus uncorrelated analog variation — evidence that Hebbian learning may operate with essentially two synaptic strengths.

Demonstrated a way to predict function from a static map of synaptic connectivity even when many neuronal and synaptic dynamics are unknown. A differentiable model of the fly visual system, constrained by full EM connectivity, had its unknown parameters tuned by gradient descent to match the known motion selectivity of a subset of neurons — showing how the strong constraints of a connectome can be leveraged even without complete dynamics.

From the MICrONS cubic-millimeter EM volume with matched in vivo calcium imaging, the authors asked whether neurons with similar response properties are preferentially connected. Normalizing for axon–dendrite proximity and using a digital twin to quantify each cell's tuning, they found that cells with similar feature tuning are more interconnected than expected, while cells with similar location tuning are less interconnected — a striking general wiring rule.

Found a linear relationship between synapse size and strength, providing the missing link in assigning physiological weights to synapses reconstructed from electron microscopy — a key step on the path to decoding information from static brain structure.

The first studies to visualize and quantify the "synaptic engram" in the hippocampus and amygdala — an essential mechanism by which memory is stored — demonstrating how memory strength correlates with the number and size of synapses among engram cells.

Revealed how the brain establishes and maintains learned sound–action associations through selective strengthening of corticostriatal synapses, and showed these patterns persist even after behavioral reversal.

Revealed the precise spatiotemporal choreography of synaptic plasticity during memory consolidation and introduced an optogenetic tool for selectively erasing LTP with unprecedented temporal precision.