Pith. sign in

REVIEW 6 minor 46 references

Feynman Diagrams

T0 review · 0 major / 6 minor · reviewed 2026-08-10 · deepseek-v4-flash

Pith's one-line read A concise pedagogical guide to Feynman diagrams, deriving the Feynman-rule dictionary from any Lagrangian and surveying modern tree- and loop-level computation methods.

desk verdict A clean, accurate review of standard material; no new physics, but a dependable teaching resource that deserves refereeing as a review article. read the letter →

arxiv 2501.08354 v1 pith:WQ52CQNK submitted 2025-01-13 hep-ph

classification hep-ph MSC 81T1881Q30
keywords Feynmandiagramsrulesperturbativequantumfieldtheoryscatteringamplitudesloopintegralsdimensionalregularizationintegration-by-partsidentitiesmultiplepolylogarithms
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 teaches the reader how Feynman diagrams arise and how to compute with them. It first shows, in a finite-dimensional Gaussian toy model, that a diagram is nothing but the pairing structure of Wick's theorem: each edge is an inverse matrix element, each internal vertex is an interaction factor, and the symmetry factor compensates for overcounting interchangeable parts. It then generalizes to relativistic quantum field theory, deriving the propagator from the bilinear part of the Lagrangian and the vertex rule from each interaction term of order three or more in the fields. The final sections survey the modern toolkit — spinor-helicity variables, colour decomposition, recurrence relations for trees, and integration-by-parts plus differential equations for loops — that lets practitioners compute amplitudes far beyond what naive diagram counting would allow.

What carries the argument

The central object is the Feynman-rule dictionary encoded in equations (20) and (32), together with the loop-number formula $l = n_{\rm int} - r_{\rm int} + 1$. The dictionary is justified by the toy model, where the propagator $(P^{-1})_{ij}$ is the contraction of two Gaussian fields and the interaction vertex $(-\lambda)$ arises from expanding $\exp\bigl(-\tfrac{\lambda}{24} \sum_i \phi_i^4\bigr)$; the same pattern carries over to the path integral once fields are valued at spacetime points. The machinery does the work of turning a diagram into a number: it assigns propagators to edges, vertices from the interaction Lagrangian, polarization factors to external lines, loop integrations to unconstrained momenta, symmetry factors, and minus signs for closed fermion loops.

What would settle it

Compute the one-loop tadpole integral in dimensional regularization both by the Feynman-rule dictionary and by direct Gaussian integration of the toy-model generating functional with a specific positive-definite matrix $P$ and coupling $\lambda$; the two answers must match exactly in the Laurent expansion in $\varepsilon$. A mismatch, or a mismatch between direct differentiation and eq. (6) for a random $P$, would show the dictionary or the boundary-term assumption fails.

Watch

Extended reading notes

Core claim

Feynman diagrams are not a mnemonic but a faithful translation of Gaussian integration: the perturbative expansion of a path integral is organized by the same combinatorial object as the differentiation of a generating functional. Given a Lagrangian, the dictionary is fixed: bilinear terms define the propagator as $i$ times the inverse of the kinetic operator, terms with three or more fields define vertices through a symmetrized Fourier-space factor, and each diagram is weighted by momentum conservation, loop integrations over unconstrained momenta, and the inverse symmetry factor. If the dictionary is accepted, computing a scattering amplitude reduces to drawing all graphs at a given order, translating each, and summing.

Load-bearing premise

The load-bearing premise is that all fields fall off fast enough at infinity that partial integrations have no boundary terms, which is used twice — in deriving the propagator from the bilinear Lagrangian and in justifying the integration-by-parts identities of Section 5.

Editorial extensions

If this is right

  • Given any perturbatively well-defined Lagrangian, one can write down the amplitude at any fixed order as a sum over diagrams; the paper gives the explicit recipe in its boxed Feynman rules.
  • Tree amplitudes can in principle be computed for any number of external particles by algebra alone; the obstacle is the factorial growth of the diagram count, which colour ordering and cyclic-ordered primitive amplitudes reduce substantially.
  • Efficient evaluation of multi-gluon tree amplitudes is achieved by off-shell recurrence relations, which reuse lower-point currents and scale polynomially, and by on-shell recursion relations, which give compact analytic formulae.
  • Loop integrals are made well-defined by dimensional regularization; tensor integrals reduce to scalar integrals, and scalar integrals reduce to a finite set of master integrals via integration-by-parts identities.
  • Master integrals satisfy a closed system of first-order differential equations; when an epsilon-factorized basis exists, the solution is written as iterated integrals, most often multiple polylogarithms.

Reading between the lines

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

  • Because the toy model already contains the full dictionary, the same derivation could be run for any theory whose kinetic operator is invertible; one extension is to test whether the boundary-term assumption breaks down for theories on manifolds with boundaries, in which case the momentum-space propagator would need surface corrections.
  • The paper notes that not all Feynman integrals are multiple polylogarithms — the two-loop sunrise integral draws on a genus-one curve — so a natural next step is to classify which Calabi-Yau geometries appear in the alphabets of epsilon-factorized differential equations.
  • The counting table implies that at very high multiplicity, Monte Carlo sampling over helicity configurations becomes comparatively more attractive, because it trades the $2^n$ prefactor against avoiding the $N_{\rm terms}^2$ cost of squaring a large sum of diagrams.
  • For teaching, the toy model suggests that Feynman diagrams can be introduced as pure Gaussian combinatorics before any quantum field theory is mentioned, making the step to path integrals smaller for students.
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.

Referee Report

0 major / 6 minor

Summary. This manuscript is a pedagogical review of Feynman diagrams. It begins with a finite-dimensional Gaussian toy model in which the expansion of e^{-λφ^4} is shown to generate the same combinatorial structures as Wick contractions, including symmetry factors. It then derives Feynman rules from a generic Lagrangian with bilinear and interaction terms, illustrates them in φ^4 theory and QED, and computes the tree-level e^-e^+ → μ^-μ^+ amplitude in Eq. (40) and a helicity amplitude in Eq. (42). The second half surveys modern methods for tree amplitudes (spinor helicity, colour decomposition, off-shell and on-shell recursion) and loop integrals (dimensional regularisation, integration-by-parts identities, differential equations, multiple polylogarithms). The paper is explicitly a selection rather than a comprehensive treatment, as acknowledged in the Conclusions.

Significance. This is a review article rather than original research, and its value lies in its pedagogical clarity rather than in new technical results. The finite-dimensional toy model in Section 2 is a genuine strength: it lets the reader verify Eq. (6) by direct Gaussian differentiation and see precisely how symmetry factors arise. The derivation of propagator and vertex Feynman rules in Section 3 is standard and internally consistent, and the worked φ^4 and QED examples are correct. The tree-level amplitude in Eq. (40) and the helicity amplitude in Eq. (42) are also correct up to the local notation issues listed below. The survey of modern tree and loop methods is current, well referenced, and the explicit scope limitation in the Conclusions is appropriately stated. There are no fitted parameters or data calibrations, so circularity is not a concern.

minor comments (6)
  1. [Eq. (3)] The prefactor of the Gaussian integral is printed with exponent n/2; it should be N/2 to match the N-dimensional integral in Eq. (1).
  2. [Eq. (40)] The gauge-dependent term of the photon propagator contains an undefined momentum q in the denominator; both denominators should be p_{12}^2. The term vanishes by the Dirac equation, so the final equality is unaffected, but the notation should be corrected.
  3. [Eq. (32)] The vertex rule presupposes that O in Eq. (29) carries the conventional symmetry prefactor for identical fields (1/4! for φ^4 and 1 for the QED vertex). If a reader symmetrizes O before applying the permutation sum, the rule double-counts; this convention should be stated explicitly.
  4. [Section 3, boundary terms] The assumption that fields fall off rapidly enough to drop boundary terms is stated, but a one-sentence caveat would help: this restricts the derivation to perturbation theory around the vacuum or a trivial background, and nontrivial backgrounds may require modified momentum-space Feynman rules.
  5. [Section 5, Eq. (49)] The vanishing of the total-derivative integral should be described as a defining property of dimensional regularisation rather than as an ordinary consequence of absent boundary terms, since the integrand of a divergent integral need not fall off sufficiently fast.
  6. [Minor typos] Please correct typographical slips such as "the the Lagrangian" in Section 3, "conceptional" in Section 5, and the redundant "therefore" in the sentence preceding the tree-level example in Section 4.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the Feynman-rule dictionary and loop-method survey are self-contained textbook derivations, with no fitted input renamed as a prediction.

full rationale

This is a pedagogical review with no fitted parameters, no empirical predictions, and no calibrated inputs. The Feynman-rule dictionary in Section 3 follows from the path-integral expansion of the action: the propagator rule in eq. (20) is defined as the inverse of the bilinear operator P extracted from the Lagrangian in eqs. (17)–(19), while the vertex rule in eq. (32) is obtained from the Fourier-transformed interaction term with the permutation sum; both are then checked against the φ^4 and QED examples. The toy model in Section 2 independently exhibits the same combinatorial structure through Gaussian differentiation, so the diagrammatic rules are not being used to predict their own input. The loop-diagram methods are surveyed with references and involve no circular reduction: the integration-by-parts identity in eq. (49) is a defining property of the dimensional regulator, and the boundary-term fall-off assumption stated in Section 3 is a normal perturbative-QFT domain condition, not a hidden fit. The self-citations ([9], [21], [44]) support peripheral computational or technical summaries and are either standard textbook material or independently checkable numerical methods; none of them carries the central claim that Feynman rules translate diagrams into integrals. The paper's own statement in the Conclusions that page limitations forced a selection is a scope limitation, not a circularity. Overall, the derivation chain is self-contained and any agreement with textbook Feynman rules is the intended derivation, not a circularity.

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

This is a review article. It introduces no free parameters fitted to data and no invented entities. The content rests entirely on standard axioms of perturbative quantum field theory: Gaussian functional integration, the legitimacy of the perturbative expansion, rapid field fall-off for partial integration, and the rules of dimensional regularization.

assumptions (5)
  • domain assumption The path integral and Gaussian source-differentiation formalism used in equations (3), (10), and (12) are well-defined, and Wick's theorem (eq. 4) gives the correct combinatorics.
    Invoked in Sections 2 and 3 as standard QFT background; the paper does not derive the measure or the validity of functional differentiation.
  • domain assumption The perturbation series in the coupling is a valid way to approximate amplitudes; the expansion in powers of lambda (eq. 2) and in powers of g (eq. 38) can be rearranged and truncated.
    The entire diagram expansion depends on treating interaction terms as small. The paper states the coupling is small but does not discuss convergence or the asymptotic nature of the series.
  • domain assumption Fields fall off rapidly at infinity so that boundary terms in partial integrations vanish.
    Explicitly stated in Section 3, page 4, and used in eq. (49) for integration-by-parts identities. If boundary terms survive, the standard momentum-space rules require modification.
  • standard math Dimensional regularization is valid: the angular integration identity (47) defines Feynman integrals for complex D, and analytic continuation allows a Laurent expansion in epsilon.
    The paper relies on this established method, citing refs [22-24], without re-deriving its mathematical foundations.
  • domain assumption Integration-by-parts identities (49) generate all linear relations among Feynman integrals, and the number of master integrals is finite.
    This is a standard result in the multiloop literature, cited as ref [36], and is used in Section 5 to set up the differential equation method.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Feynman Diagrams." pith.science (2026). https://pith.science/paper/WQ52CQNK

@misc{pith2026250108354,
  author       = {Pith},
  title        = {Pith review of: Feynman Diagrams},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WQ52CQNK}},
  note         = {Machine review of arXiv:2501.08354}
}
read the original abstract

We give a concise and pedagogical introduction to Feynman diagrams. After discussing a toy model which requires only undergraduate mathematics, we focus on relativistic quantum field theory. We review the derivation of Feynman rules from the Lagrangian of the theory and we discuss modern methods to compute tree and loop diagrams.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

46 extracted references · 14 canonical work pages

  1. [1]

    R. P . Feynman, Space-time approach to nonrelativistic q uantum mechanics, Rev. Mod. Phys. 20 (1948) 367–387, doi: 10.1103/RevModPhys. 20.367

  2. [2]

    G. C. Wick, The Evaluation of the Collision Matrix, Phys. Rev. 80 (1950) 268–272, doi: 10.1103/PhysRev.80.268

  3. [3]

    M. E. Peskin, D. V . Schroeder, An Introduction to Quantum Field Theory , Perseus Books 1995

  4. [4]

    M. D. Schwartz, Quantum Field Theory and the Standard Mod el, Cambridge University Press 2014

  5. [5]

    Srednicki, Quantum field theory, Cambridge Universit y Press 2007

    M. Srednicki, Quantum field theory, Cambridge Universit y Press 2007

  6. [6]

    Mangano, Stephen J

    Michelangelo L. Mangano, Stephen J. Parke, Multiparton amplitudes in gauge theories, Phys. Rept. 200 (1991) 301–36 7, doi: 10.1016/ 0370-1573(91)90091-Y , hep-th/0509223

  7. [7]

    Dixon, Calculating scattering amplitudes effic iently, in: Theoretical Advanced Study Institute in Elemen tary Particle Physics (T ASI 95): QCD and Beyond 1996, pp

    Lance J. Dixon, Calculating scattering amplitudes effic iently, in: Theoretical Advanced Study Institute in Elemen tary Particle Physics (T ASI 95): QCD and Beyond 1996, pp. 539–584, hep-ph/9601359

  8. [8]

    Henriette Elvang, Y u-tin Huang, Scattering Amplitudes , Cambridge University Press 2015, 1308.1697

Show all 46 references
  1. [9]

    Stefan Weinzierl, T ales of 1001 Gluons, Phys. Rept. 676 ( 2017) 1–101, doi: 10.1016/j.physrep.2017.01.004, 1610.05318

  2. [10]

    Notes Phys

    Simon Badger, Johannes Henn, Jan Christoph Plefka, Sim one Zoia, Scattering Amplitudes in Quantum Field Theory, Le ct. Notes Phys. 1021 (2024) pp., doi: 10.1007/978-3-031-46987-9 , 2306.05976

  3. [11]

    De Causmaecker, R

    P . De Causmaecker, R. Gastmans, W. T roost, T ai Tsun Wu, M ultiple Bremsstrahlung in Gauge Theories at High-Energies . 1. General Formalism for Quantum Electrodynamics, Nucl. Phys. B 206 (1 982) 53–60, doi: 10.1016/0550-3213(82)90488-6

  4. [12]

    Zhan Xu, Da-Hua Zhang, Lee Chang, Helicity Amplitudes f or Multiple Bremsstrahlung in Massless Nonabelian Gauge Th eories, Nucl. Phys. B 291 (1987) 392–428, doi: 10.1016/0550-3213(87)90479-2

  5. [13]

    Petros Draggiotis, Ronald H. P . Kleiss, Costas G. Papad opoulos, On the computation of multigluon amplitudes, Phys . Lett. B 439 (1998) 157–164, doi: 10.1016/S0370-2693(98)01015-6 , hep-ph/9807207

  6. [14]

    Cvitanovic, P

    P . Cvitanovic, P . G. Lauwers, P . N. Scharbach, Gauge Inv ariance Structure of Quantum Chromodynamics, Nucl. Phys. B 186 (1981) 165–186, doi: 10.1016/0550-3213(81)90098-5

  7. [15]

    Berends, W

    Frits A. Berends, W. Giele, The Six Gluon Process as an Ex ample of Weyl-Van Der Waerden Spinor Calculus, Nucl. Phys. B 294 (1987) 700–732, doi: 10.1016/0550-3213(87)90604-3

  8. [16]

    Mangano, Stephen J

    Michelangelo L. Mangano, Stephen J. Parke, Zhan Xu, Dua lity and Multi - Gluon Scattering, Nucl. Phys. B 298 (1988) 65 3–672, doi: 10.1016/0550-3213(88)90001-6

  9. [17]

    Berends, W

    Frits A. Berends, W. T . Giele, Recursive Calculations f or Processes with n Gluons, Nucl. Phys. B 306 (1988) 759–808, doi:10.1016/ 0550-3213(88)90442-7 . Feynman Diagrams 11

  10. [18]

    Ruth Britto, Freddy Cachazo, Bo Feng, New recursion rel ations for tree amplitudes of gluons, Nucl. Phys. B 715 (2005 ) 499–522, doi: 10.1016/j.nuclphysb.2005.02.030, hep-th/0412308

  11. [19]

    Ruth Britto, Freddy Cachazo, Bo Feng, Edward Witten, Di rect proof of tree-level recursion relation in Y ang-Mills t heory, Phys. Rev. Lett. 94 (2005) 181602, doi: 10.1103/PhysRevLett.94.181602, hep-th/0501052

  12. [20]

    Smirnov, Analytic tools for Feynman integr als, Springer T racts Mod

    Vladimir A. Smirnov, Analytic tools for Feynman integr als, Springer T racts Mod. Phys. 250 (2012) 1–296, doi: 10.1007/978-3-642-34886-0

  13. [21]

    A Comprehensive T reatment for Students and Researchers, UNITEXT for Physic s, Springer 2022, doi:10.1007/978-3-030-99558-4 , 2201.03593

    Stefan Weinzierl, Feynman Integrals. A Comprehensive T reatment for Students and Researchers, UNITEXT for Physic s, Springer 2022, doi:10.1007/978-3-030-99558-4 , 2201.03593

  14. [22]

    Gerard ’t Hooft, M. J. G. Veltman, Regularization and Re normalization of Gauge Fields, Nucl. Phys. B 44 (1972) 189–2 13, doi: 10.1016/ 0550-3213(72)90279-9

  15. [23]

    C. G. Bollini, J. J. Giambiagi, Dimensional Renormaliz ation: The Number of Dimensions as a Regularizing Parameter , Nuovo Cim. B 12 (1972) 20–26, doi: 10.1007/BF02895558

  16. [24]

    G. M. Cicuta, E. Montaldi, Analytic renormalization vi a continuous space dimension, Lett. Nuovo Cim. 4 (1972) 329– 332, doi: 10.1007/ BF02756527

  17. [25]

    Kinoshita, Mass singularities of Feynman amplitude s, J

    T . Kinoshita, Mass singularities of Feynman amplitude s, J. Math. Phys. 3 (1962) 650–677, doi: 10.1063/1.1724268

  18. [26]

    T . D. Lee, M. Nauenberg, Degenerate Systems and Mass Sin gularities, Phys. Rev. 133 (1964) B1549–B1562, doi: 10.1103/PhysRev.133. B1549

  19. [27]

    O. V . T arasov, Connection between Feynman integrals ha ving different values of the space-time dimension, Phys. Re v. D 54 (1996) 6479– 6490, doi:10.1103/PhysRevD.54.6479, hep-th/9606018

  20. [28]

    O. V . T arasov, Generalized recurrence relations for two loop propagator integrals with arbitrary masses, Nucl. Ph ys. B 502 (1997) 455–482, doi:10.1016/S0550-3213(97)00376-3 , hep-ph/9703319

  21. [29]

    A. V . Kotikov, Differential equations method: New tech nique for massive Feynman diagrams calculation, Phys. Lett . B 254 (1991) 158–164, doi:10.1016/0370-2693(91)90413-K

  22. [30]

    A. V . Kotikov, Differential equation method: The Calcu lation of N point Feynman diagrams, Phys. Lett. B 267 (1991) 1 23–127, doi: 10.1016/ 0370-2693(91)90536-Y , [Erratum: Phys.Lett.B 295, 409–409 (1992)]

  23. [31]

    A 110 (1997) 1435–1452, doi: 10.1007/BF03185566, hep-th/9711188

    Ettore Remiddi, Differential equations for Feynman gr aph amplitudes, Nuovo Cim. A 110 (1997) 1435–1452, doi: 10.1007/BF03185566, hep-th/9711188

  24. [32]

    Gehrmann, E

    T . Gehrmann, E. Remiddi, Differential equations for tw o-loop four-point functions, Nucl. Phys. B 580 (2000) 485–5 18, doi: 10.1016/ S0550-3213(00)00223-6 , hep-ph/9912329

  25. [33]

    F . V . Tkachov, A Theorem on Analytical Calculability of Four Loop Renormalization Group Functions, Phys. Lett. B 10 0 (1981) 65–68, doi:10.1016/0370-2693(81)90288-4

  26. [34]

    K. G. Chetyrkin, F . V . Tkachov, Integration by Parts: Th e Algorithm to Calculate beta Functions in 4 Loops, Nucl. Phy s. B 192 (1981) 159–204, doi: 10.1016/0550-3213(81)90199-1

  27. [35]

    Laporta, High-precision calculation of multiloop F eynman integrals by difference equations, Int

    S. Laporta, High-precision calculation of multiloop F eynman integrals by difference equations, Int. J. Mod. Phys . A 15 (2000) 5087–5159, doi:10.1142/S0217751X00002159, hep-ph/0102033

  28. [36]

    A. V . Smirnov, A. V . Petukhov, The Number of Master Integ rals is Finite, Lett. Math. Phys. 97 (2011) 37–44, doi: 10.1007/s11005-010-0450-0 , 1004.4199

  29. [37]

    Xiao Liu, Y an-Qing Ma, Chen-Y u Wang, A Systematic and Ef ficient Method to Compute Multi-loop Master Integrals, Phys . Lett. B 779 (2018) 353–357, doi: 10.1016/j.physletb.2018.02.026, 1711.09572

  30. [38]

    Zhi-Feng Liu, Y an-Qing Ma, Determining Feynman Integr als with Only Input from Linear Algebra, Phys. Rev. Lett. 129 (22) (2022) 222001, doi:10.1103/PhysRevLett.129.222001, 2201.11637

  31. [39]

    Xiao Liu, Y an-Qing Ma, AMFlow: A Mathematica package fo r Feynman integrals computation via auxiliary mass flow, Com put. Phys. Commun. 283 (2023) 108565, doi: 10.1016/j.cpc.2022.108565, 2201.11669

  32. [40]

    Henn, Multiloop integrals in dimensional r egularization made simple, Phys

    Johannes M. Henn, Multiloop integrals in dimensional r egularization made simple, Phys. Rev. Lett. 110 (2013) 2516 01, doi: 10.1103/ PhysRevLett.110.251601, 1304.1806

  33. [41]

    Kuo-Tsai Chen, Iterated Path Integrals, Bull. Amer. Ma th. Soc. 83 (1977) 831–879

  34. [42]

    A. B. Goncharov, Multiple polylogarithms, cyclotomy a nd modular complexes, Math. Res. Lett. 5 (1998) 497

  35. [43]

    Borwein, David M

    Jonathan M. Borwein, David M. Bradley , David J. Broadhu rst, Petr Lisonek, Special values of multiple polylogarith ms, T rans. Am. Math. Soc. 353 (2001) 907–941, doi: 10.1090/S0002-9947-00-02616-7 , math/9910045

  36. [44]

    Jens Vollinga, Stefan Weinzierl, Numerical evaluatio n of multiple polylogarithms, Comput. Phys. Commun. 167 (20 05) 177, doi: 10.1016/j. cpc.2004.12.009, hep-ph/0410259

  37. [45]

    Brown, Multiple zeta values and periods of moduli spa ces M0,n, C

    F . Brown, Multiple zeta values and periods of moduli spa ces M0,n, C. R. Acad. Sci. Paris 342 (2006) 949

  38. [46]

    Bourjaily , et al., Functions Beyond Multiple P olylogarithms for Precision Collider Physics, in: Snowmas s 2021, 2203.07088

    Jacob L. Bourjaily , et al., Functions Beyond Multiple P olylogarithms for Precision Collider Physics, in: Snowmas s 2021, 2203.07088

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.