QFT Research Atlas

Research is where stable QFT background is used to assess live questions. The atlas does not turn a recent paper into a verdict: it separates established results from assumptions, identifies correlations between apparently independent claims, and states what observation, proof, or benchmark would change the assessment.

Evidence cutoff. The initial atlas synthesis includes sources available through 11 August 2026. No immutable Research edition has yet been released; cite the primary literature linked on each page.

Choose a research task

If you want to…	Start with	What you will find
orient yourself in an area	Fields	programs, characteristic observables, entry paths, benchmarks, and durable obstructions
examine one unsettled proposition	Frontier questions	established facts, competing formulations, negative evidence, and resolution criteria
decide how a calculation can bear on a question	Methods	applicability assumptions, error budgets, cross-checks, and failure modes
check what a finite body of results supports	Evidence briefs	source-by-source comparisons with shared inputs and important exclusions exposed
cite a frozen Research assessment	Research editions	immutable snapshots when reviewed publication records exist

The five routes answer different questions. A field guide is an entry map, not a ranking. A frontier dossier assesses a precisely bounded question, not a whole subject. A method map compares inferential reach, not prestige. An evidence brief tests a claim against a declared source set, not all conceivable literature. An edition is a frozen record, not the current mutable assessment.

How to read an assessment

Every detailed page begins by fixing the theory, dimension, regime, observable, assumptions, and exclusions that control its conclusion. Its direct assessment should be read together with four nearby elements:

1. the cutoff, which limits how current the assessment can be;
2. the background path, which points to the stable concepts used rather than rederiving them;
3. the dependence analysis, which prevents shared data, code, priors, or perturbative inputs from being counted twice; and
4. the change criteria, which state what could genuinely strengthen, weaken, or resolve the conclusion.

Terms such as established, supported, contested, and open are always scope-qualified. Agreement across two analyses is strong evidence only to the extent that their decisive inputs and failure modes are independent. Conversely, disagreement need not imply new physics: it can diagnose an unmodeled systematic, a mismatched observable, or a comparison outside the domain where both calculations are controlled.

Authored inventory at this cutoff

The atlas contains twelve field guides, twenty-four frontier dossiers, ten method maps, and eight evidence briefs. The evidence briefs presently cover the muon anomalous magnetic moment, the $W$-boson mass, the three-dimensional Ising bootstrap benchmark, the finite-temperature QCD equation of state, hydrodynamic attractors, Planckian dissipation, analogue Hawking radiation, and the scope of the quantum null energy condition.

These counts describe the authored route inventory, not current citable coverage. Current citable Research coverage is empty because no immutable edition has yet been released. The selection deliberately mixes experimentally active comparisons, nonperturbative numerical benchmarks, formal inequalities, and cross-method claims so that the atlas exposes different kinds of uncertainty.

Consequential omissions at this cutoff include dedicated evidence briefs on electroweak vacuum metastability, conformal-window determinations, and real-time quantum simulation; those subjects appear only as field, question, or method pages. The atlas is not a systematic review of all QFT research, and absence from it is not a judgment about importance.

Boundaries of the atlas

Research pages do not replace the Volumes’ stable explanations, reproduce news feeds, or assign authority by citation count. They do not conceal unresolved disagreements inside a single average, and they do not translate a successful calculation in a benchmark system into evidence for every nearby theory. Where primary results conflict, the conflict remains visible.

The assessment pages may change as errors are corrected or evidence is reassessed. The Research editions directory explains the additional preservation and review records required before an assessment can be cited as an immutable atlas snapshot.

Keeping the assessment current

Field guides and method maps should be reconsidered at least annually, frontier dossiers and evidence briefs at least every six months, and briefs tied to active official combinations or measurements every three months. A new result triggers earlier review when it changes a central value or uncertainty, supplies an independent test, invalidates a source, or changes the domain in which a method is controlled.