Evidence Briefs

An evidence brief asks what a declared, finite source set supports about one bounded claim. Each brief states its cutoff, keeps incompatible results visible, traces shared inputs, and distinguishes the observation from the interpretation placed on it.

Evidence cutoff. The briefs below were synthesized from primary and authoritative sources available through 11 August 2026. None has yet been frozen into a citable Research edition; cite the primary literature on the relevant brief.

Evidence sets in this atlas

Brief	Bounded assessment	Main reason for caution
Muon anomalous magnetic moment	the final experimental value is precise, while the size of any Standard Model discrepancy depends on the hadronic-vacuum-polarization evaluation	lattice and dispersive inputs are not yet mutually consistent
$W$-boson mass	recent LHC measurements and the electroweak fit agree within uncertainties, while the precise CDF II value remains incompatible	combining results requires common treatment of PDFs, recoil, calibration, and correlations
Three-dimensional Ising bootstrap benchmark	crossing and unitarity isolate exceptionally small allowed regions under explicit assumptions	finite numerical truncation and spectral assumptions delimit what is certified
Finite-temperature QCD equation of state	independent lattice programs agree on the zero-density crossover equation of state over their shared continuum-controlled range	action, scale-setting, charm, and high-temperature regimes are not identical
Hydrodynamic attractor	attractor-like loss of initial-condition sensitivity appears across several controlled theoretical descriptions	evidence from heavy-ion data is indirect and model dependent
Planckian dissipation	order-$k_B T/\hbar$ relaxation scales recur in many materials	inferred times depend on transport models and do not establish a universal microscopic bound
Analogue Hawking radiation	analogue horizons robustly exhibit mode conversion and, in some platforms, partner correlations	these experiments test analogue kinematics, not astrophysical black-hole evaporation
Quantum null energy condition scope	QNEC has strong flat-space QFT proofs and broad checks	curved-background statements require additional geometric and renormalization hypotheses

How the matrices are constructed

The evidence relation in each row is deliberately typed: a source may support the bounded claim, count against it, qualify its scope, replicate part of it, or provide context only. “Independent” is not a synonym for “different paper.” Shared collaborations, datasets, calibration chains, simulation ensembles, perturbative inputs, software, and theoretical priors are noted whenever they reduce the effective number of tests.

An uncertainty quoted by a source is retained with its definition. The briefs do not manufacture a combined significance when covariance is unavailable or when the inputs measure materially different quantities. A null result constrains only the parameter and regime it actually probes; a successful benchmark establishes neither universal correctness nor unique interpretation.

Using a brief responsibly

Read the direct assessment and its principal limitation together. Then inspect:

1. whether the observable matches the claim;
2. whether the dominant uncertainties are statistical, systematic, theoretical, or structural;
3. whether apparently agreeing sources share decisive inputs;
4. which serious contrary result remains unresolved; and
5. what future result would change the assessment.

For the methods behind a comparison, use the method maps. For the wider unresolved problem, follow the linked frontier dossier. Stable definitions and derivations remain in the linked Volumes.

Reassessment cadence

Evidence briefs should be reconsidered within six months. The muon and $W$-mass briefs, which depend on active experimental and theory combinations, should be checked within three months or immediately after a new official result. Corrections, retractions, covariance releases, and independently reproduced calculations trigger review even when the headline central value is unchanged.