Observed Galaxy Abundances at z > 6 Exceed Halo-Limited Predictions in COSMOS-Web

AbstractRecent observations with the James Webb Space Telescope have revealed a substantial population of massive galaxies at redshifts ( z \gtrsim 6 ), posing a persistent challenge to standard (\Lambda)CDM predictions of early structure formation. In this work, we present a systematic, model-agnostic analysis of galaxy number densities in the COSMOS2025 field ((0.54,\mathrm{deg}^2)), focusing on stellar mass–selected samples with (\log(M_\star/M_\odot) \geq 9) over the redshift range (5.5 \leq z \leq 11).We find that the observed number densities exceed baseline (\Lambda)CDM expectations by factors of (5!-!20\times), increasing monotonically with redshift. This excess persists under conservative stress tests, including stellar mass shifts of up to (+0.5) dex, photometric redshift contamination fractions up to 50%, and (3\sigma) cosmic variance corrections. Even in the most extreme combined scenario, the tension remains at the (\gtrsim 3\times) level across all redshift bins.Importantly, the result is driven by independent observables: the total galaxy counts, their redshift evolution, and the monotonic increase in specific star formation rate with redshift. No tuning of the initial mass function, selection thresholds, or survey geometry is able to reconcile the observations with standard halo abundance predictions.We interpret these findings as evidence for a regime-dependent breakdown of equilibrium-based structure formation assumptions at early cosmic times. While we refrain from proposing a specific alternative cosmological model, we show that the data naturally motivate a transition-regime framework in which early galaxies form under conditions not captured by conventional collapse-limited growth. The empirical result itself is robust and falsifiable, and places strong constraints on any viable theory of early galaxy formation.