Conjecture Atlas
Frozen public edition v0.5.1. Read the current version of this page.

The main objective

For a smooth, projective and geometrically integral variety X/FX/F of dimension nn over a number field, construct a canonical pairing

hX:CH⁡r(X)Q,hom×CH⁡n+1−r(X)Q,hom⟶R. h_X:\operatorname{CH}^r(X)_{\mathbb Q,\mathrm{hom}}\times\operatorname{CH}^{n+1-r}(X)_{\mathbb Q,\mathrm{hom}}\longrightarrow\mathbb R.

The local conjecture supplies Green forms; a global construction must also establish the local sum, independence of choices and rational equivalence. The paper states each of these requirements explicitly.

The logical implications

There are two routes to the global pairing: enlarge the obstruction-zero domain, or construct it directly from strong Green forms with the additional analytic hypotheses.

HypothesesConclusionArgument
A Green pair on the prescribed complement and a global actual primitive of its curvature.A zero-curvature Green form for that prescribed cycle, against all actual tests.Curvature correction
The height construction on the intersection of the local obstruction kernels, and vanishing for global classes in both complementary degrees.The same pairing is defined on both full homologically trivial Chow groups.Enlargement of the domain
Strong G in both degrees, together with the hypotheses on proper intersections, principal divisors, reciprocity and good reduction in the direct construction.A symmetric Chow pairing with the stated base-extension and biextension properties. Correspondence adjunction and classical comparison have separate proofs.Direct construction of the pairing

The research problems

Steps 1–3 concern existence. Step 4 identifies the resulting height. Steps 5–6 ask for additional arithmetic properties; they do not follow from existence alone.

1. Green forms and finite trace

Construct a Green pair for every prescribed cycle over the original non-Archimedean field.

D[GZ]+δZ=[ΩZ]D[G_Z]+\delta_Z=[\Omega_Z], where D=d′d′′D=d'd'' and GZ∈Br−1,r−1(Xan∖∣Z∣an)G_Z\in\mathcal B^{r-1,r-1}(X^{\mathrm{an}}\setminus|Z|^{\mathrm{an}}).

Proposed argument
Prove the supported construction and its trace identities on the actual form space. Trace must preserve differential relations and adjunction against every compact actual test, with the full field-degree normalization.
Consequence
Once the curvature has a global actual primitive, subtracting it gives Strong G on the same support complement.
Prerequisites
The actual differential calculus and the supported Chern-form construction.
Precise statement and research plan

2. Vanishing of the Bott–Chern class

Show that homologically trivial cycles have zero curvature class in the actual Bott–Chern quotient.

oX,Kr(CH⁡r(X)Q,hom)=0\mathfrak o^r_{X,K}\bigl(\operatorname{CH}^r(X)_{\mathbb Q,\mathrm{hom}}\bigr)=0.

Proposed argument
First apply the algebraic-equivalence and Chow-correspondence arguments on their stated domains. For the remaining classes, seek a comparison theorem showing that vanishing of the cohomological cycle class forces every actual obstruction period to vanish.
Consequence
Together with Step 1, this is a sufficient route to Strong G. The tropical approach is an alternative method for constructing the required actual primitives, with its own descent and integration comparisons.
Prerequisites
Step 1 for the passage from obstruction vanishing to a Green form.
Precise statement and research plan

3. Construction of the global height pairing

Establish a local sum which is independent of Green representatives and descends through rational equivalence in both variables.

hX(Z,W)=∑v[Fv:Qv][F:Q] λv(Zv,Wv)h_X(Z,W)=\sum_v \dfrac{[F_v:\mathbb Q_v]}{[F:\mathbb Q]}\,\lambda_v(Z_v,W_v).

Proposed argument
Use the existing construction on the intersection of the obstruction kernels at all finite completions, then prove vanishing on localizations of global classes in both degrees. Alternatively, use directly supplied strong Green forms with proper-intersection formulas, compatibility with principal divisors under pushforward, reciprocity and vanishing at good places.
Consequence
A symmetric pairing on the desired Chow groups, with the specified field normalization. A finite-detector description requires its own comparison theorem.
Prerequisites
The local calculus and global reciprocity. The second route assumes Strong G directly and need not prove vanishing of the chosen obstruction map.
Precise statement and research plan

4. Correspondence adjunction and classical heights

Prove that the constructed pairing has the required correspondence identity and agrees with the classical height on common domains.

hX(Γ∗z,w)=hY(z,tΓ∗w)h_X(\Gamma_*z,w)=h_Y(z,{}^t\Gamma_*w).

Proposed argument
Apply the proper-intersection projection formula to correspondences. Compare the finite-place terms on supplied admissible arithmetic models, and match the archimedean determinant and Poincaré normalizations. At bad reduction, construct the required vertical corrections.
Consequence
Identification with the classical Beilinson–Bloch pairing on the common domain, together with the normalization needed for subsequent height identities.
Prerequisites
Step 3 and the additional projection-formula and admissible-model hypotheses used by each comparison.
Precise statement and research plan

5. Arithmetic Lefschetz and positivity

For pointed products of curves, prove the Chow vanishing and central positivity assertions used by the arithmetic reduction theorem.

Π∗CH⁡c(Y)Q,hom=0(k≠2c−1),Y=∏i=1kCi\Pi_*\operatorname{CH}^c(Y)_{\mathbb Q,\mathrm{hom}}=0\quad(k\ne2c-1),\qquad Y=\prod_{i=1}^{k}C_i, where Π=⨂i(ΔCi−oi×Ci−Ci×oi)\Pi=\bigotimes_i(\Delta_{C_i}-o_i\times C_i-C_i\times o_i) is the projector onto the product of the h1(Ci)h^1(C_i) factors.

Proposed argument
Work separately on off-central vanishing, injectivity of the central algebraic Abel–Jacobi map, and real positive definiteness of the residual height. Keep the fixed product polarization and the height axioms H1–H4 throughout.
Consequence
Under H1–H4, these assertions yield arithmetic hard Lefschetz, primitive positivity and complementary nondegeneracy by the central-block identities.
Prerequisites
The height system H1–H4 on all required subproducts, source curves of correspondences and finite extensions; these arithmetic assertions are additional to height existence.
Precise statement and research plan

6. Integral Northcott finiteness

Prove bounded-height finiteness for the projected images of integral cycles, first over a fixed field and then in bounded degree.

#{z∈Λ≤d:q(z)≤B}<∞\#\{z\in\Lambda_{\le d}:q(z)\le B\}<\infty. Here qq is the signed residual quadratic height and Λ≤d\Lambda_{\le d} consists of projected integral classes defined over extensions of degree at most dd of the fixed base field.

Proposed argument
Over each fixed field, combine real positivity with a common finite-rank lattice, or use real positivity and a proved height gap with a finite Kummer quotient for some integer m ≥ 2. Then bound ramification of the minimal fields of definition of bounded-height classes.
Consequence
A common finite-rank lattice and real positivity give fixed-field Northcott; the additional ramification control gives bounded-degree Northcott. Positive definiteness alone does not give either finiteness assertion.
Prerequisites
The residual quadratic height and integral image from the arithmetic programme. Integral control and field control are separate requirements.
Precise statement and research plan

Formal statements and available proofs

The programme paper gives the precise objectives and conditional arguments. The progress page collects the constructions already available and their remaining hypotheses.