The radius valley is one of the most robust features in exoplanet demographics. Around sun-like stars and early M dwarfs, planet sizes cluster into two populations — smaller super-Earths below about 1.5 Earth radii and larger sub-Neptunes above about 2 Earth radii — with a conspicuous gap between them. The gap is explained by atmospheric loss: planets in the valley had their hydrogen envelopes stripped by photoevaporation or core-powered mass loss, leaving bare rocky cores below and gas-retaining planets above. The valley is a boundary between two outcomes of the same formation process.
Gillis, Cloutier, and Pass (2602.23364) surveyed 8,134 mid-to-late M dwarfs observed by TESS and found 77 planet candidates. The radius valley is gone. The planet size distribution is unimodal, peaking at 1.25 Earth radii, with super-Earths outnumbering sub-Neptunes 5.5 to 1. The bimodal structure that defines planetary demographics around larger stars does not exist around smaller ones.
The disappearance has an explanation: if planets around low-mass stars form primarily through water-rich pebble accretion rather than gas accretion, there are simply fewer sub-Neptunes to create the upper peak. The valley vanishes not because the stripping mechanism stops working but because the population it would divide was never bimodal in the first place. There is no gap because there was never a second peak.
The general principle: a feature observed across one parameter range can disappear when the range extends, not because the mechanism fails but because the conditions that created the feature were themselves parameter-dependent. The radius valley is not a universal boundary between rocky and gaseous planets. It is a consequence of how planets form around stars of a particular mass range. Change the stellar mass enough, and the formation pathway changes, and the feature that seemed fundamental turns out to be conditional.