{"id":"ac17b1ed-432d-4173-b74d-da8dba834e5b","arxiv_id":"1908.03346","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"All known 3d N=8 and N=6 SCFT moduli spaces are quotients by reflection groups, predicting two new N=8 theories and proving the equivalence of two ABJM theories.","lead":"Tachikawa and Zafrir show that the moduli spaces of all known 3d theories with maximal or near-maximal supersymmetry, after suitable finite gaugings, are quotients by real or complex reflection groups. The pattern predicts two new maximally supersymmetric theories and proves an unexpected equivalence between two ABJM Chern-Simons theories.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The H3/H4 prediction rests on an unproved bijection: for N=8 the unique 'oldest relative' and the 'relatives iff same Γ' premise are inferred from examples, not derived.","rationale":"The reader's conditional verdict correctly identifies the weakest point. The paper's concrete calculations are careful and provide real support: Sec. 3 gives explicit moduli-space computations for the known ABJM/ABJ/BLG families, and Sec. 4 supplies both a path-integral argument and an explicit superconformal-index match for the equivalence of two ABJM formulations. These are not where the argument is fragile. The fragility is structural: the periodic-table claim and the H3/H4 prediction require that the reflection group of the locally oldest relative is a unique, complete label for N=8 theories, and that statement is inferred from a finite list of examples and a web of partly conjectural dualities, not proved. The paper is honest about this, even flagging a possible 4d counterexample that would kill the N=6 version, but honesty does not remove the gap. I therefore agree with the reader: a conditional verdict is appropriate, and the concrete test above targets the specific unproved duality that would most directly falsify or support the one-to-one premise.","tokens_in":30728,"tokens_out":15831,"duration_ms":175855,"concrete_test":"Test the least-supported duality needed for the 'relatives iff same Γ' premise, the G2 row of Table 1: compute the superconformal index of the BLG-type theory (SU(2)_6 × SU(2)_-6)/Z2 via standard Chern-Simons-matter localization, and compute the index of the G2 SYM IR fixed point from a UV completion with manifest N=8 or from the 4d N=3 S-fold reduction chain described in Sec. 2.2.3. Compare the two expansions order by order. If they disagree at any order, the one-to-one correspondence already fails at I2(6) = WG2 and the H3/H4 inference loses its evidential basis; if they agree, the unique-label premise gains one of its missing links.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Sec. 1.3 first states in full generality that 'we have not eliminated the possibility that Q can have more than one locally oldest relative whose reflection groups are different,' then asserts for N=8 that 'the inspection of the list of known N=4 theories and various data computed for them' reveals a unique oldest relative and that two N=8 theories are relatives iff their reflection groups agree. This is the load-bearing step. The H3/H4 prediction in Sec. 1.1 follows only if (i) the locally oldest relative is unique for every N=8 theory, so Γ is a well-defined label, and (ii) the existing list of N=8 SCFTs is complete enough that the only unused real reflection groups are H3 and H4. Neither is derived. The supporting Table 1 itself contains conjectural identifications: the G2 row requires the proposed duality between (SU(2)_6 × SU(2)_-6)/Z2 and the low-energy limit of G2 SYM (Sec. 2.2.3), which the authors present as indirect evidence, not as an established equivalence. If a second oldest relative with a different Γ exists, the label is not unique; if an as-yet-unknown N=8 theory has an orbifold group that is not a real reflection group, the gap labeled H3/H4 need not be filled by hidden theories. The explicit moduli computations and the Sec. 4 index check are good support, but they do not establish the bijection. On the N=6 side the paper explicitly disclaims uniqueness, so the predictive force of the paper rests entirely on the N=8 uniqueness/completeness premise.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes that the moduli spaces of all known 3d N=8 and N=6 SCFTs, after suitable gaugings of finite symmetries so that one passes to a 'locally oldest' relative, take the form C^{4r}/Γ, where Γ is a real reflection group for N=8 and a complex reflection group for N=6. The known N=8 examples realize the dihedral and Weyl groups; the only remaining irreducible real reflection groups are H_3 and H_4, which the paper suggests correspond to two yet-to-be-discovered N=8 theories. The known N=6 theories realize the infinite family G(k,p,N), and the paper suggests looking for N=6 theories associated with exceptional complex reflection groups. Along the way, the paper gives a detailed derivation of the moduli spaces for (SU(N)_k × SU(N)_-k)/Z_m ABJM theories, (U(N+x)_k × U(N)_-k)/Z_p ABJ(M) theories, and U Sp(2N)_k × SO(2)_-2k theories, including checks of the consistency of the finite quotients with the Chern-Simons levels. It also proves, by a path-integral argument and a superconformal index check, the equivalence of (U(N)_k × U(N)_-k)/Z_k and (SU(N)_k × SU(N)_-k)/Z_N.","tokens_in":30947,"tokens_out":10862,"duration_ms":116766,"significance":"The paper's concrete computations are a genuine contribution: the moduli-space derivations in Sec. 3 are explicit and internally consistent, the quotient consistency checks via instanton numbers and monopole charges are nontrivial, and the proof of the ABJM equivalence in Sec. 4, supported by the superconformal index comparison in Sec. 4.3, appears sound. The organization of known 3d N>=6 SCFTs by real and complex reflection groups is a clean and potentially useful organizing principle, and the specific H_3/H_4 prediction is falsifiable in principle. The authors are appropriately explicit that the H_3/H_4 gap is a conjecture rather than a theorem, and the paper does not hide the fact that the N=6 side lacks a uniqueness statement. If the reflection-group classification can be placed on a firmer basis, this would be an important step toward a periodic table of highly supersymmetric 3d theories.","major_comments":[{"comment":"The H_3/H_4 prediction and the 'periodic table' of N=8 theories rest on two unproved assertions: that every N=8 theory has a unique locally oldest relative, and that two N=8 theories are relatives if and only if their reflection groups agree. The paper presents these as consequences of 'the inspection of the list of known N=4 theories and various data', but no systematic derivation or exhaustive check is given. This is load-bearing because, as the paper itself acknowledges in the preceding paragraph, a theory can have more than one locally oldest relative with different reflection groups; if that happens for some N=8 theory, or if an unknown N=8 theory has a moduli space not of reflection-group form, the H_3/H_4 gaps need not be filled by hidden theories. I ask the authors to either provide a systematic verification of uniqueness and completeness over the known examples, or state explicitly in the abstract and in Sec. 1.1 that the H_3/H_4 conjecture is conditional on these two premises.","section":"Sec. 1.3, paragraph beginning 'For N=8, however...'"},{"comment":"Table 1 is used as evidence that the known N=8 theories are in one-to-one correspondence with real reflection groups, but at least one row is itself conjectural: the G_2 row requires identifying (SU(2)_6 × SU(2)_-6)/Z_2 with the low-energy limit of G_2 SYM, an equivalence that Sec. 2.2.3 supports only through indirect 4d S-fold evidence. Since the one-to-one correspondence claim depends on every real reflection group being realized by exactly one known theory, this entry should be marked as conjectural in Table 1, and the table should distinguish established dualities from proposed ones.","section":"Sec. 2.2.3 and Table 1"}],"minor_comments":[{"comment":"The phrase 'the inspection of the list of known N=4 theories' is confusing; since the surrounding discussion is about N=8 theories, it should either read 'known N=8 theories' or explicitly explain that N=8 theories are being viewed as a subclass of 3d N=4 theories.","section":"Sec. 1.3, first paragraph after the N=8 uniqueness assertion"},{"comment":"Reference [42] is listed as 'O. Bergman to appear'; this is incomplete. Please provide a preprint number or a more specific citation, or remove it and cite the relevant published work.","section":"Reference [42]"},{"comment":"The statement that real reflection groups are dihedral, Weyl, or H_3,H_4 should specify 'irreducible real reflection groups', since direct products of real reflection groups are also real reflection groups and appear in some of the moduli spaces discussed in the paper.","section":"Sec. 1.3, list of real reflection groups"}],"recommendation":"minor_revision","confidential_remarks":"The paper is a solid contribution with explicit and mostly convincing computations; the main weakness is that the headline H_3/H_4 prediction is logically conditional on an unproved uniqueness/completeness premise. I believe this can be addressed by a careful revision that either strengthens the evidence for the premise or clearly labels the prediction as conjectural in the abstract. The ABJM equivalence proof and the moduli-space computations do not need major changes."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know: this paper gives a periodic-table proposal for 3d N=8 and N=6 SCFTs using real and complex reflection groups, and proves a new duality between two ABJM versions. The moduli-space computations are careful, the path-integral equivalence in Sec. 4 is convincing, and the index check makes it very plausible. This is not a paper of vague analogies; the concrete machinery is doing real work.\n\nWhat is actually new: the identification of all known N=6 moduli spaces with quotients by G(k,p,N) is a useful unification, and the H3/H4 prediction is a sharp, falsifiable claim. The proof that (SU(N)_k × SU(N)_{-k})/Z_N equals (U(N)_k × U(N)_{-k})/Z_k for all N,k goes beyond earlier coprime checks and is the cleanest result in the paper. Credit is also due for the explicit consistency checks of finite quotients with Chern-Simons levels—this is exactly the kind of detail that separates a real derivation from a hand-wave.\n\nThe soft spot is precisely what the stress-test note says: the N=8 bijection—unique oldest relative, and relatives iff same reflection group—is inferred from examples, not proved, and the H3/H4 prediction depends on that plus a completeness assumption about the known N=8 list. The authors are honest about this, and they even flag a 4d rank-2 example that would disprove the broader scheme, which is more than most speculative classifications do. The G2 row in Table 1 also leans on an unproved duality (BLG k=6 vs. G2 SYM), though that is presented as indirect evidence rather than established. None of this sinks the paper; it means the central organizing claim is a conjecture with strong supporting evidence, not a theorem.\n\nWho should read it: anyone working on 3d supersymmetric dualities, moduli spaces, or M2-brane theories. The ABJM equivalence alone is worth the read, and the reflection-group classification is a useful lens even if the strong bijection fails. A serious referee should engage with it—the open questions are well posed and the computations are checkable. I would send it out, and I would expect the revision to sharpen the status of the uniqueness premise.","headline":"A genuinely useful organizing claim for 3d N≥6 SCFTs, with solid moduli-space computations and a clean ABJM equivalence, but the H3/H4 prediction rests on an unproved uniqueness/completeness premise that the authors mostly acknowledge.","tokens_in":31619,"tokens_out":957,"would_cite":true,"duration_ms":12865,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"After suitable finite gaugings, every known highly supersymmetric 3d theory has a moduli space shaped by a reflection group, and two sporadic symmetries predict new N=8 theories.","keywords":["3d N=8 SCFTs","3d N=6 SCFTs","moduli spaces","real reflection groups","complex reflection groups","ABJM theory","ABJ theory","finite gaugings"],"falsifier":"Construct or discover a 3d N=8 SCFT whose moduli space, after gauging every non-anomalous finite symmetry, is not $\\mathbb{C}^{4r}/\\Gamma$ with $\\Gamma$ a real reflection group; or prove rigorously that no 3d N=8 SCFT can have moduli space $\\mathbb{C}^4/H_4$, in which case the proposed one-to-one labeling by real reflection groups is false.","tokens_in":30430,"feed_emoji":"🔬","tokens_out":9310,"duration_ms":85176,"temperature":0.7,"pith_summary":"The paper sets out to show that three-dimensional SCFTs with maximal or near-maximal supersymmetry are usefully labelled by reflection groups: after gauging suitable finite symmetry groups, the moduli space of every known N=8 theory is $C^{4r}/\\Gamma$ with $\\Gamma$ a real reflection group, and the moduli space of every known N=6 theory is $C^{4r}/\\Gamma$ with $\\Gamma$ a complex reflection group. Because real reflection groups are classified as dihedral groups, Weyl groups, or the two sporadic groups $H_3$ and $H_4$, and because the dihedral and Weyl cases are already realized by the BLG and super Yang-Mills theories, the paper concludes that two N=8 SCFTs for $H_3$ and $H_4$ have yet to be discovered. For N=6, all known examples fall into the infinite family $G(k,p,N)$, leaving the exceptional complex reflection groups as a natural place to look for new theories. Along the way, the paper proves that the ABJM theories $(SU(N)_k\\times SU(N)_{-k})/\\mathbb{Z}_N$ and $(U(N)_k\\times U(N)_{-k})/\\mathbb{Z}_k$ are the same SCFT.","feed_headline":"Icosahedral symmetries predict two undiscovered 3d N=8 theories","feed_subtitle":"Known 3d N=8 and N=6 moduli spaces all become reflection-group quotients; two sporadic groups lack theories.","key_machinery":"The load-bearing object is the orbifold moduli space $\\mathbb{C}^{4r}/\\Gamma$ read through the Chevalley-Shephard-Todd theorem: the invariant ring is a polynomial ring precisely when $\\Gamma$ is a complex reflection group, so the degrees of the invariant generators are a fingerprint of the theory. The paper determines $\\Gamma$ by isolating the subgroup of the unbroken gauge group that acts trivially on both the matter fields and the monopole operators; the Chern-Simons level dictates which finite quotients are consistent, and the surviving identifications---phase rotations $z_i\\mapsto e^{2\\pi i p/k}z_i$ and pairwise rotations $(z_i,z_j)\\mapsto(e^{2\\pi i/k}z_i,e^{-2\\pi i/k}z_j)$ together with permutations---generate exactly $G(k,p,N)$, reducing to dihedral and Weyl groups in the N=8 cases. The proof that the two ABJM variants agree is carried by a one-form $U(1)$ symmetry: integrating along this symmetry direction produces a delta function that identifies the two determinant $U(1)$s, converting $(U(N)_k\\times U(N)_{-k})/\\mathbb{Z}_k$ into $(SU(N)_k\\times SU(N)_{-k})/\\mathbb{Z}_N$.","core_discovery":"The central claim, stated on the paper's own terms, is that reflection groups give a classification scheme for 3d $\\mathcal{N}\\ge 6$ SCFTs rather than just a bookkeeping device. Supersymmetry forces the moduli space of an N=8 theory to be $\\mathbb{R}^{8r}/\\Gamma$ and that of an N=6 theory to be $\\mathbb{C}^{4r}/\\Gamma$, with $\\Gamma$ induced from an action that commutes with the R-symmetry; the new observation is that, once finite gaugings are used to pass to a `locally oldest' relative, $\\Gamma$ is always a reflection group. The authors then argue that for N=8 the group $\\Gamma$ appears to distinguish relatives: two N=8 theories are related by finite gaugings exactly when their $\\Gamma$s agree. Since the classification of real reflection groups is short, this turns the zoo of known theories into a periodic table whose empty cells are $H_3$ and $H_4$, and it identifies the missing N=8 theories as the main open problem. For N=6, the same analysis places every known theory in the family $G(k,p,N)$ and motivates a search for theories labelled by exceptional complex reflection groups.","pith_inferences":["A practical consequence the paper leaves implicit: if $H_3$ and $H_4$ theories exist, their sphere partition functions and superconformal indices should be determined by the reflection-group data alone, for instance through the degrees of the basic invariants, so one could test the prediction without possessing a Lagrangian.","The uniqueness of the oldest relative could fail in a way that is now well defined: an N=8 theory with two locally oldest relatives carrying different reflection groups, or two non-relative theories with the same $\\Gamma$, would overturn the one-to-one labeling while leaving the moduli-space observations intact.","The one-form-symmetry integration used to prove the ABJM equivalence looks like a general duality machine, not a special trick; applying it to other Chern-Simons-matter theories with several abelian factors may produce further unexpected identifications among theories with different-looking gauge groups."],"forward_implications":["If the periodic table is right, the N=8 SCFTs are labelled by the finite list of real reflection groups; the empty entries $H_3$ and $H_4$ are concrete predictions of two new theories.","The equivalence $(U(N)_k\\times U(N)_{-k})/\\mathbb{Z}_k = (SU(N)_k\\times SU(N)_{-k})/\\mathbb{Z}_N$ holds for all $N$ and $k$, removing the coprime restriction that limited earlier checks of the same duality.","A 3d N=6 theory whose reflection group is one of the exceptional groups $G_4$ through $G_{37}$ would be genuinely new and cannot be a relative of any known ABJ(M) theory.","The scheme sharpens the contrast with 4d: there the crystallographic condition selects Weyl groups for N=4 and crystallographic complex groups for N=3, while the non-crystallographic groups $H_3$, $H_4$, and the non-crystallographic exceptionals appear to be special to 3d."],"supporting_citations":[{"why":"Defines the $U(N)_k\\times U(N)_{-k}$ ABJM Chern-Simons-matter theories that form the core N=6 family.","marker":"[3]"},{"why":"Defines the ABJ theories $U(N+x)_k\\times U(N)_{-k}$ and the orthosymplectic variants that provide the remaining known N=6 examples.","marker":"[4]"},{"why":"Classifies all Lagrangian N=6 theories up to gauge algebra, giving the exhaustive list of known N=6 moduli spaces analyzed in Section 3.5.","marker":"[16]"},{"why":"Supplies the dualities between enhanced ABJ(M) theories and super Yang-Mills theories of types A, B, C and D on which the N=8 table depends.","marker":"[19]"},{"why":"Earlier comparison of $U(N)\\times U(N)$ and $SU(N)\\times SU(N)$ moduli spaces that Section 4 extends to all $N$ and $k$.","marker":"[20]"},{"why":"Establishes the dualities used to identify the BLG theories at $k=2$ and $k=4$ with other N=8 SCFTs.","marker":"[21]"},{"why":"Establishes the duality identifying the $k=3$ BLG theory with the interacting $SU(3)/\\mathbb{Z}_3$ super Yang-Mills theory.","marker":"[22]"},{"why":"Provides the abelian duality by which the path integral over a one-form symmetry direction yields the delta function used to prove the ABJM equivalence.","marker":"[40]"},{"why":"Supplies the generalized global symmetry framework used to define relatives, to justify gauging finite 1-form symmetries, and to check anomaly consistency of Chern-Simons levels.","marker":"[30]"}],"fun_headline_variants":["Reflection groups predict two missing N=8 superconformal theories","Icosahedral symmetry hints at two new 3d N=8 theories","Sporadic groups H3 and H4 point to undiscovered N=8 SCFTs","All known 3d N≥6 SCFTs fit reflection group classification","Two ABJM theories shown equivalent by reflection group analysis"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The periodic table rests on the premise, inferred from known examples and conjectural dualities rather than proved, that every N=8 theory has a unique `oldest' relative and that two N=8 theories are relatives exactly when their reflection groups agree; if that fails, the $H_3$ and $H_4$ prediction loses its basis.","fun_headline_variants_meta":{"raw":{"variants":["Reflection groups predict two missing N=8 superconformal theories","Icosahedral symmetry hints at two new 3d N=8 theories","Sporadic groups H3 and H4 point to undiscovered N=8 SCFTs","All known 3d N≥6 SCFTs fit reflection group classification","Two ABJM theories shown equivalent by reflection group analysis"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000737,"raw_usage":{"total_tokens":3340,"prompt_tokens":1036,"completion_tokens":2304,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":652,"completion_tokens_details":{"reasoning_tokens":2206}},"tokens_in":652,"tokens_out":2304,"duration_ms":15596,"temperature":1.0,"reasoning_tokens":2206,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:16:29.618377+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct or discover a 3d N=8 SCFT whose moduli space, after gauging every non-anomalous finite symmetry, is not $\\mathbb{C}^{4r}/\\Gamma$ with $\\Gamma$ a real reflection group; or prove rigorously that no 3d N=8 SCFT can have moduli space $\\mathbb{C}^4/H_4$, in which case the proposed one-to-one labeling by real reflection groups is false.","supporting_citations":[{"cited_title":"ABCD of 3d ${\\cal N}=8$ and 4 Superconformal Field Theories","cited_arxiv_id":"1108.3647","evidence_quote":"Supplies the dualities between enhanced ABJ(M) theories and super Yang-Mills theories of types A, B, C and D on which the N=8 table depends."},{"cited_title":"Relating U(N)xU(N) to SU(N)xSU(N) Chern-Simons Membrane theories","cited_arxiv_id":"1001.4779","evidence_quote":"Earlier comparison of $U(N)\\times U(N)$ and $SU(N)\\times SU(N)$ moduli spaces that Section 4 extends to all $N$ and $k$."},{"cited_title":"New and old N=8 superconformal field theories in three dimensions","cited_arxiv_id":"1103.3548","evidence_quote":"Establishes the dualities used to identify the BLG theories at $k=2$ and $k=4$ with other N=8 SCFTs."},{"cited_title":"A New Duality Between $\\mathcal{N}=8$ Superconformal Field Theories in Three Dimensions","cited_arxiv_id":"1708.07861","evidence_quote":"Establishes the duality identifying the $k=3$ BLG theory with the interacting $SU(3)/\\mathbb{Z}_3$ super Yang-Mills theory."}],"review_version":1}