The asymptotically-free gauge theories

Ben Gripaios; Khoi Le Nguyen Nguyen

REVIEW 1 major objections 5 minor 2 cited by

The asymptotically-free gauge theories

T0 review · 1 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper proves that dimension and Dynkin index increase with every Dynkin label, making the classification of asymptotically-free fermionic gauge theories finite, explicit, and complete.

desk verdict Solid classification paper with a correct monotonicity proof and useful tables, but the completeness of the exceptional-algebra table rests on an undocumented scan that referees should ask to see. read the letter →

arxiv 2507.12348 v1 pith:CUYI5DKA submitted 2025-07-16 hep-th hep-ph

Ben Gripaios , Khoi Le Nguyen Nguyen This is my paper

classification hep-thhep-ph

keywords asymptoticfreedomgaugetheoriesfermionicmattersimpleLiealgebrasirreduciblerepresentationsDynkinindexlabelsbetafunction

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper undertakes the first complete classification of asymptotically-free gauge theories in four dimensions with purely fermionic matter. It proves a new monotonicity theorem: both the dimension and the Dynkin index of any irreducible representation of a simple Lie algebra increase strictly when any Dynkin label is increased. Since asymptotic freedom is just a bound on the Dynkin index, this makes the search finite and produces explicit tables of all non-trivial asymptotically-free irreducible representations. The tables show that no such representation uses more than two nonzero Dynkin labels and no label exceeds four. The same method extends to scalar and supersymmetric matter by changing coefficients in the beta function.

What carries the argument

The argument is carried by the one-loop $\beta$-function coefficient for a simple gauge factor, proportional to $-22T(\lambda_{\mathrm{adj}})+4T(\lambda_f)$, so asymptotic freedom for fermionic matter is the bound $T(\lambda)<(11/2)T(\lambda_{\mathrm{adj}})$. The paper's new monotonicity theorem says that increasing any Dynkin label strictly increases both the Weyl-dimension $D(\lambda)$ and the Dynkin index $T(\lambda)$, so once the bound is crossed in any coordinate no further increment can restore asymptotic freedom. The tables are generated by checking the finitely many highest weights left after this bound, using Weyl's dimension formula and Dynkin's formula for $T(\lambda)$.

What would settle it

Independently enumerate every highest weight of $G_2$, $F_4$, $E_6$, $E_7$, and $E_8$ with $T(\lambda)<(11/2)T(\lambda_{\mathrm{adj}})$ and compare with Table V; a single missing weight would disprove the completeness claim.

Watch

Extended reading notes

Core claim

The central claim is that Tables I-V list exactly the non-trivial asymptotically-free irreducible representations of every simple Lie algebra, for purely fermionic matter in four dimensions. The tables follow from a new theorem: both $D(\lambda)$ and $T(\lambda)$ are strictly increasing functions of each Dynkin label $m_i$ in $\lambda=\sum_i m_i\omega_i$. Because asymptotic freedom for a simple summand is the inequality $T(\lambda)<(11/2)T(\lambda_{\mathrm{adj}})$, monotonicity leaves only finitely many weights to check, and the finite check gives the tables. In particular, the tables show that at most two Dynkin labels can be non-zero and no single label can exceed four.

Load-bearing premise

The completeness of Tables I-V depends on an unshown brute-force scan over the exceptional algebras; if that scan missed any candidate weight, the central classification would be incomplete.

Editorial extensions

If this is right

For any simple Lie algebra, there are only finitely many non-trivial asymptotically-free irreducible representations, and Tables I-V give the complete list.
In every such representation at most two Dynkin labels are non-zero and no label is larger than four, sharply restricting the fermion content of any asymptotically-free model.
For a semisimple gauge algebra, a representation is asymptotically free exactly when each simple summand satisfies the same index bound, so reducible asymptotically-free spectra can be enumerated summand by summand using the tables.
For $A_n$, anomaly cancellation becomes a finite linear-integer problem; the paper gives 10,036 asymptotically-free and anomaly-free reducible representations for $n=4$.
Changing the coefficients in the beta function extends the same classification procedure to scalar matter and to supersymmetric theories.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Not developed in the paper, the same monotonicity argument could produce classifications for other one-loop thresholds, such as a lower bound that selects near-conformal or walking theories, simply by replacing $(11/2)T(\lambda_{\mathrm{adj}})$ with another constant.
The only unexhibited step, the brute-force check for the exceptional algebras, could be made fully reproducible by releasing the code or the intermediate counts; an independent recomputation would settle Table V without changing any physics.
The anomaly coefficients in Table I turn the search for chiral asymptotically-free $A_n$ theories into a linear Diophantine problem, so a complete database for all $n$ is computationally within reach.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The genuinely new and useful thing here is the proof that both dimension and Dynkin index are strictly increasing in each Dynkin label, plus the complete tables for AF irreducible reps of all simple Lie algebras. The monotonicity argument is short, elementary, and correct; it converts an infinite check into a finite one and yields the clean structural result that at most two labels can be nonzero and none exceed four. The tables also explicitly correct omissions in two earlier partial lists, which is a real service to the model-building community.

The soft spot is exactly where the stress-test lands: Table V, for the exceptional algebras, is obtained by a 'sledgehammer' with no code, no intermediate counts, and no bounds on the Dynkin labels that were scanned. The monotonicity theorem guarantees finiteness but does not by itself tell you which weights survive the bound T(lambda) < (11/2)T(adj). A reader cannot verify that the enumeration was exhaustive without either the code or a more careful analytic argument. I did a couple of spot checks (e.g., the 64 of G2 and the 273 of F4 both fail the bound, as the stress-test says) and found no actual counterexample, but that is not the same as checking the whole space. This is a verifiability gap, not a demonstrated error. A referee should ask the authors to deposit the scan script or give the per-weight bounds M_i and candidate counts.

Footnote [5] is a second, minor soft spot: it asserts, without proof, a classification of AF and anomaly-free product reps for An, settling a question from an earlier paper. That is a side remark and does not affect the main tables, but if it is meant to be a claim, it needs a proof or a reference to one.

Overall the central argument holds up. The reader's conditional verdict is fair, and my own confidence is moderate for the same reasons. This is a paper for gauge-theory model builders and for people interested in Dynkin-index inequalities; it does not change conceptual foundations but it is a solid, usable reference. I would send it to peer review, with the request that the authors make the exceptional-algebra enumeration reproducible.

Referee Report

1 major / 5 minor

Summary. The paper studies asymptotic freedom in four-dimensional gauge theories with purely fermionic matter, reducing the problem to the Dynkin index T(λ) of the matter representation. The key new result is a proof that for every simple Lie algebra, both the dimension D(λ) and the Dynkin index T(λ) are strictly increasing functions of each Dynkin label. Combined with the one-loop beta-function condition T(f) < (11/2)T(adj), this implies that for each simple algebra there are finitely many non-trivial asymptotically-free irreducible representations. The authors provide Tables I–V listing these representations for An, Bn, Cn, Dn and the five exceptional algebras, including dimensions, Dynkin indices, and anomaly information for An, and they describe how to extend the results to reducible and semisimple representations. They also identify two irreducible theories with exactly vanishing one-loop beta function and discuss anomaly-free chiral reducible representations.

Significance. If correct, this is a useful reference classification of the irreducible building blocks of asymptotically-free gauge theories with fermionic matter. The monotonicity proof is simple, self-contained, and correct, and it converts an infinite enumeration problem into a finite one; the classical tables follow from standard index formulas and the monotonicity theorem. There are no fitted parameters and the central derivation uses only Weyl's and Dynkin's formulas, so I see no circularity. The main weakness is that the completeness of Table V for the exceptional algebras is not independently verifiable from the text. I found no evidence that the tables are numerically wrong, but the undocumented 'sledgehammer' scan is a genuine reproducibility gap that needs to be closed before the central completeness claim can be fully accepted.

major comments (1)

[Table V / 'We now outline how the Tables are obtained'] Completeness of Table V is not verifiable as written. The text says only: 'For the exceptional algebras, there are anyway only a finite number of cases to check, so we hit them with a sledgehammer.' Since Table V is part of the central claim that the non-trivial AF irreducible representations of each simple algebra are exactly those in Tables I–V, the reader needs to be able to check that the scan covered every dominant weight λ with T(λ) < (11/2)T(λ_adj). Please provide at least one of: (i) explicit upper bounds M_i on each Dynkin label for each exceptional algebra, obtained from monotonicity and the inequality T(M_i ω_i) < (11/2)T(λ_adj); (ii) the number of candidate dominant weights considered in each rank and the number retained; or (iii) a short, self-contained script or pseudocode that reproduces Table V. This is a load-bearing step, not a cosmetic issue.

minor comments (5)

[Footnote [5]] The assertion that the only AF and anomaly-free chiral gauge theory that is a product of irreducible representations of type An has n=6 and fermion representation ω2 ⊗ ω6 is nontrivial and is stated without proof or citation. Since it is not needed for the main classification, please either provide a derivation or clearly mark it as a separate computational result with supporting material.
[Abstract / Tables I–V] The abstract and title could be read as promising a complete list of asymptotically-free gauge theories, whereas the tables list only irreducible representations; reducible representations are obtained by the described algorithm and are not enumerated except for examples. Please state explicitly that the tables classify irreducible representations and that reducible theories are generated by the given finite procedure.
[Table I caption and rows] The notation in Table I for the n ranges uses merged symbols such as '{1, 2∗..., 15∗}' and the asterisk/† markers are explained only indirectly in the caption. Please clarify which values of n carry each marker, or use separate notation for 'omitted in [8]' and 'omitted in [9]'.
[Section 'We finish by proving...'] In the proof of strict monotonicity of D(λ), the sentence 'By the definition of ω_i, each such shift is nonnegative' is correct but would benefit from one explanatory clause: each positive root is a nonnegative integer combination of simple roots, so (ω_i, α_j) ≥ 0 for every positive root α_j.
[Page 4] Typo: 'coresponding' should be 'corresponding' in the sentence about the SO(14) theory with fermion rep ω3.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the classification follows from a proven monotonicity theorem plus Weyl's and Dynkin's formulas; the only self-citations are peripheral and the exceptional-algebra scan is a verifiability gap, not a circular step.

full rationale

The derivation chain is self-contained. The target Tables I-V are obtained from Weyl's dimension formula and Dynkin's index formula, the one-loop beta-function condition T(lambda) < (11/2)T(adj), and the paper's own proof (Section 'We finish by proving...') that D(lambda) and T(lambda) are strictly increasing in each Dynkin label. That theorem is proven from the manifestly positive shifts (omega_i, alpha_j) and (omega_i, omega_i) + 2(delta, omega_i) + 2(lambda, omega_i), not assumed. Finiteness and the truncation to a finite candidate set follow from the theorem; the remaining checks are arithmetic evaluations of the same formulas. No parameter is fitted and no 'prediction' is a renamed input. The only self-citations (refs. [15] and [17]) appear in footnotes and concern peripheral side results about product reps and infinitely many chiral anomaly-free reps; the central classification does not depend on them. The undocumented exceptional-algebra 'sledgehammer' scan is a reproducibility/verifiability gap, not a circular step: an unshown exhaustive computation is not the same as assuming the conclusion. Therefore no circular reduction can be exhibited.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

No free parameters or invented entities. The classification is computed from standard Lie algebra formulas and the one-loop beta function; there is no fitting to data. The main external assumptions are the one-loop AF criterion, Weyl/Dynkin formulas, and the anomaly formula of [7]. The completeness of the exceptional-algebra tables relies on an unshown brute-force computation.

assumptions (5)

domain assumption A simple gauge group factor is asymptotically free iff its one-loop beta coefficient satisfies 4T(f) < 22T(adj) for Weyl fermions.
Used throughout: Section (iii), the algorithm paragraph after fact (vi), and the Tables define AF by this inequality. The authors rely on the one-loop beta function as the criterion for asymptotic freedom.
standard math The irreducible representations of simple Lie algebras are labelled by Dynkin labels, with dimensions and Dynkin indices given by Weyl's and Dynkin's formulas.
Invoked in Section (iv) and used to compute all table entries; the formulas are standard results from Lie representation theory.
standard math Every semisimple Lie algebra decomposes as a direct sum of simple algebras, and a representation is a direct sum of irreducible representations.
Fact (i) and (ii) in the introduction; needed to reduce the classification to simple summands.
standard math For a product of representations, T(λ⊗λ') = D(λ)T(λ') + D(λ')T(λ), and an irreducible of a direct product is a tensor product of irreducibles of the factors.
Used in the semisimple reduction paragraph and in footnote [5] to compute Dynkin index contributions from multiple gauge factors.
domain assumption The anomaly coefficient for A_n irreps is computed by the Banks-Georgi formula cited as [7].
Used to fill the A column in Table I and to solve anomaly cancellation for reducible reps; the formula is taken from the cited literature without proof.

how reviews work

0 comments

Cite this review

Pith. "Pith review of The asymptotically-free gauge theories." pith.science (2026). https://pith.science/paper/CUYI5DKA

@misc{pith2026250712348,
  author       = {Pith},
  title        = {Pith review of: The asymptotically-free gauge theories},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CUYI5DKA}},
  note         = {Machine review of arXiv:2507.12348}
}

read the original abstract

We show how to classify the asymptotically-free gauge theories in four spacetime dimensions, focussing here on the case of purely fermionic matter. The classification depends on the fact (which we prove) that both the dimension and Dynkin index of irreducible representations of a simple Lie algebra are strictly increasing functions of each Dynkin label. This implies not only that the number of asymptotically-free representations of any one semisimple Lie algebra is finite, but also that they can be written down in a systematic fashion using tables for the asymptotically-free irreducible representations of simple Lie algebras, which we supply. These tables show that at most two out of a possible ten Dynkin labels can be non-zero and that no Dynkin label can exceed four. The extension to bosonic matter or supersymmetric theories is straightforward.

Discussion (0). Sign in to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

Anomaly cancellation for two $U(1)$ factors
hep-th 2026-07 accept novelty 7.0 of 10

Abelian anomaly cancellation for rank-K U(1) summands equates to finding (K-1)-planes on a cubic hypersurface over Q; for K=2 and six fermions this is the Fano surface of the Segre cubic, whose rational components ful...
Comment to "The asymptotically-free gauge theories"
hep-th 2025-07 accept novelty 2.0 of 10

A comment showing that the Gripaios-Nguyen classification of asymptotically-free gauge theories misses anomaly cancellation constraints and prior classifications, and overstates its novelty.

Reference graph

Works this paper leans on

18 extracted references · 15 canonical work pages · cited by 2 Pith papers

[5]

Crucially, fact (v) implies that for each possible gauge Lie algebra, there are only finitely-many AF reps, once we exclude the trivial rep

and the latter are given by Dynkin’s for- mula T (λ) = D(λ) D(λadj) [(λ + δ, λ+ δ) − (δ, δ)], where 2δ is the sum of the positive roots and where a rep’s dimension is given by Weyl’s formula D(λ) = Q j (λ+δ,αj )Q j (δ,αj ) , with the products taken over the positive roots; (v) T (λ) and D(λ) are, as we will show at the end, strictly-increasing functions o...

arXiv 2025
[1]

A partial list of AF irreducible reps of type An was given without proof in [8], but omits the 15 irreducible reps indicated with a ∗ in Table I; a list of AF chiral irreducible reps of type An and AF pseudoreal irreducible reps of Cn was given without proof in [9], but omits the irreducible rep indicated with a † in Table I

work page
[2]

For the known facts, details can be found in [13, 14]

work page
[3]

This differs from the usual normalization in the physics literature, but is obviously preferable because it results in an integer-valued Dynkin index

work page
[4]

E. B. Dynkin, Trans. Am. Math. Soc. Ser. 2 6, 111 (1957)

work page 1957
[6]

For a product of reps, we instead have the for- mula T (λ ⊗ λ′) = D(λ)T (λ′) + D(λ′)T (λ); using this, together with the fact that a product of irreducible reps contains as a summand the irreducible rep whose highest weight is the sum of the highest weights of the factors, it is possible to show that the only AF and anomaly-free chiral gauge theory that i...

work page
[7]

Our ordering convention follows that of [16]

work page
[8]

Banks and H

J. Banks and H. Georgi, Phys. Rev. D14, 1159 (1976)

work page 1976

Show all 18 references

[9]

Eichten, K

E. Eichten, K. Kang, and I. Koh, J. Math. Phys.23, 2529 (1982)

1982
[10]

Cacciapaglia, K

G. Cacciapaglia, K. Kollias, A. Deandrea, and F. San- nino, (2025), arXiv:2507.06368

2025 arXiv
[11]

W. E. Caswell, Phys. Rev. Lett.33, 244 (1974)

1974
[12]

As [17] shows, there are infinitely many chiral anomaly- free irreducible reps for each n ≥ 4, but none are AF
[13]

Kumar and M

S. Kumar and M. S. Narasimhan, Math. Ann. 308, 155 (1997), arXiv:alg-geom/9511012

1997 arXiv
[14]

J. E. Humphreys, Introduction to Lie algebras and rep- resentation theory, Graduate texts in mathematics ; 9 (Springer, 1972)

1972
[15]

Freudenthal and H

H. Freudenthal and H. de Vries, Linear Lie groups, Pure and applied mathematics ; 35 (Academic Press, 1969)

1969
[16]

Gripaios and K

B. Gripaios and K. Le Nguyen Nguyen, JHEP, To appear (2025), arXiv:2501.09860

2025 arXiv
[17]

Slansky, Phys

R. Slansky, Phys. Rept. 79, 1 (1981)

1981
[18]

Gripaios and K

B. Gripaios and K. Le Nguyen Nguyen, JHEP 12, 041, arXiv:2409.15430

arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [5]

Crucially, fact (v) implies that for each possible gauge Lie algebra, there are only finitely-many AF reps, once we exclude the trivial rep

and the latter are given by Dynkin’s for- mula T (λ) = D(λ) D(λadj) [(λ + δ, λ+ δ) − (δ, δ)], where 2δ is the sum of the positive roots and where a rep’s dimension is given by Weyl’s formula D(λ) = Q j (λ+δ,αj )Q j (δ,αj ) , with the products taken over the positive roots; (v) T (λ) and D(λ) are, as we will show at the end, strictly-increasing functions o...

arXiv 2025

[2] [1]

A partial list of AF irreducible reps of type An was given without proof in [8], but omits the 15 irreducible reps indicated with a ∗ in Table I; a list of AF chiral irreducible reps of type An and AF pseudoreal irreducible reps of Cn was given without proof in [9], but omits the irreducible rep indicated with a † in Table I

work page

[3] [2]

For the known facts, details can be found in [13, 14]

work page

[4] [3]

This differs from the usual normalization in the physics literature, but is obviously preferable because it results in an integer-valued Dynkin index

work page

[5] [4]

E. B. Dynkin, Trans. Am. Math. Soc. Ser. 2 6, 111 (1957)

work page 1957

[6] [6]

For a product of reps, we instead have the for- mula T (λ ⊗ λ′) = D(λ)T (λ′) + D(λ′)T (λ); using this, together with the fact that a product of irreducible reps contains as a summand the irreducible rep whose highest weight is the sum of the highest weights of the factors, it is possible to show that the only AF and anomaly-free chiral gauge theory that i...

work page

[7] [7]

Our ordering convention follows that of [16]

work page

[8] [8]

Banks and H

J. Banks and H. Georgi, Phys. Rev. D14, 1159 (1976)

work page 1976