Author note (2026-11-02). The paper cited here, arXiv:2603.10207 (“Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events”), is by Vincent Fiorino, Sean D. Lawley, Alan E. Lindsay. The name(s) this essay attributes it to (Yvinec) do not appear among the listed authors, so the attribution below is wrong or refers to someone else's earlier work; I have not worked out which. Found by a mechanical check of each essay's named authors against the arXiv author list.
The earliest binding events carry the most directional information.
A single cell sits in a dilute chemical gradient. Molecules of chemoattractant drift through the medium, and occasionally one binds to a receptor on the cell surface. At ultra-low concentrations, binding events are rare and stochastic. The classical question: how many binding events does the cell need to determine which direction the source is?
Yvinec et al. (arXiv:2603.10207) show that the answer depends on when the bindings happen, not just how many. Early binding events — the first molecules to arrive — carry disproportionately more information about source direction than later arrivals. The mathematics is extreme-value statistics: the first arrivals are drawn from the tail of the first-passage-time distribution, and that tail is shaped by the geometry of the source-receptor configuration.
Later binding events are progressively less informative because they come from a distribution that has already been averaged over the diffusive dynamics. The first molecule to bind traveled a path that was, probabilistically, more direct. As time passes, the arriving molecules have had more opportunity to diffuse laterally, smearing the directional signal.
The consequence: a cell that weights early bindings more heavily than late ones can extract directional information from just a handful of molecular encounters. The optimal sensing strategy isn't “collect more data” — it's “pay attention to the beginning.”
This inverts the usual statistical intuition that more samples always help. In this system, the marginal information per sample decreases with time. The signal is front-loaded. A cell that listened equally to all binding events would dilute the early signal with later noise. The asymmetry in arrival times is the information, and it's consumed as the process unfolds.
Yvinec et al., "Single-cell directional sensing at ultra-low chemoattractant concentrations from extreme first-passage events," arXiv:2603.10207 (2026).