{"id":"258763a2-7ee6-4f8f-8f87-615130e7e304","arxiv_id":"1908.03816","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Aut(T_{n,r}) is exactly the bi-synchronizing transducer subgroup T B_{n,r} that preserves the cyclic order, and Out(T_{n,r}) contains a copy of Thompson's group F for n ≥ 3.","lead":"This paper determines the automorphism groups of the generalized Thompson groups T_{n,r} in terms of bi-synchronizing transducers that preserve the cyclic order. It also shows the outer automorphism groups are infinite for n ≥ 3 and that all T_{n,r} have the R-infinity property.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Main theorem rests on unproven Proposition 4.13 from unpublished BCMNO [3]; a counterexample would invalidate the bi-synchronizing characterization.","rationale":"The reader's weakest assumption identifies the same dependency on Proposition 4.13 of the unpublished BCMNO preprint. I inspected other parts of the proof for internal gaps: the reverse direction of Theorem 5.3 is densely written but can be repaired, because in a product representing the identity the accessible core states are forced to be identity by the equation Uξ = V h(ξ) with V = U; the induction in Corollary 4.16 is salvageable by applying the induction hypothesis to the shorter word νξ; and the signature homomorphism in Section 7 is plausible because the image is a group under multiplication and inverse images provide inverses. The main unresolved risk is external: without Proposition 4.13, the finite-local-action step collapses, taking the bi-synchronizing characterization with it. This is not an attack on the author; it is a call for a published proof or independent verification. The CONDITIONAL verdict is appropriate and should remain unchanged.","tokens_in":53075,"tokens_out":44619,"duration_ms":459467,"concrete_test":"Verify Proposition 4.13 by attempting to construct a counterexample: search the space of small bi-synchronizing (or just finite) transducers over C_{2,1} representing homeomorphisms that preserve ∼_t (e.g., the odometer, its powers, and products with prefix exchanges) and check, for some pair (τ,η), whether h_{τχ} ≠ h_{ηχ} for all χ up to length 10. Also check the published literature and the current version of [3] (arXiv:1605.09302) for a proof of Prop 4.13; if it is absent, the paper's main theorem should be marked conditional.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 4's route to a finite transducer for normalizer elements passes through Corollary 4.14 and Corollary 4.16, both of which import Proposition 4.13 from the unpublished BCMNO preprint [3]. Proposition 4.13 asserts that if a homeomorphism h of C_{n,r} has h_{τχ} ≠ h_{ηχ} for every χ ∈ X_n^*, then h fails to preserve the tail equivalence relation ∼_t. The contrapositive is used to produce, for each node-distance residue i, a pair (τ,η) with equal local actions; without this, Lemma 4.15 cannot conclude 'almost same fashion' for all pairs of cones, and Corollary 4.16's finite-local-action claim fails. Since Theorem 5.3 (and hence Theorem 5.4) depends on finite local actions to build a finite minimal transducer and then the synchronizing and bi-synchronizing properties, the entire automorphism characterization would need reworking if Prop 4.13 is false. No independent proof or published source is cited; [3] was unpublished at the time of this posting. This is a verifiability and correctness risk, not an internal inconsistency.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper characterizes the automorphism group of the generalized Thompson group T_{n,r} as the subgroup of the rational group R_{n,r} consisting of elements represented by finite, initial, bi-synchronizing transducers that preserve the cyclic order relation ≃ on the Cantor space (Theorem 5.4). It defines a corresponding subgroup T O_{n,r} of the transducer group O_{n,r} from Bleak–Cameron–Maissel–Navas–Olukoya (BCMNO), and proves that Out(T_{n,r}) is isomorphic to T O_{n,r}. The paper then establishes the nesting structure T O_{n,1} ⊴ T O_{n,r} ⊴ T O_{n,n-1}, equality T O_{n,r} = T O_{n,gcd(n-1,r)}, a negative answer to the BCMNO question on isomorphism of Out(G_{n,r}) versus gcd, the existence of a copy of Thompson's group F in Out(T_{n,r}) for n ≥ 3, and the R∞ property for all T_{n,r}. The proofs use transducer dynamics, viable combinations, and explicit transducer constructions.","tokens_in":53336,"tokens_out":10454,"duration_ms":88097,"significance":"Assuming the validity of the imported Proposition 4.13 from [3], the main theorem is a natural and substantial extension of the BCMNO characterization of Aut(G_{n,r}) to the T_{n,r} family. The structural results for Out(T_{n,r}) (infinite, containing F, lattice relations, R∞ property) were previously known only for special cases, and the negative answer to the gcd question is a concrete new contribution. The paper is detailed and provides explicit transducers for key examples, including the element of T O_4 that is not in any T O_{4,r} for r < 3, and the generators of the F-subgroup. The main results are coherent and mostly self-contained except for the reliance on [3]. If Proposition 4.13 is correct, the paper meets the standard for acceptance; the remaining issues are local corrections.","major_comments":[{"comment":"Proposition 4.13 is stated without proof and is attributed to the unpublished preprint [3] (arXiv:1605.09302). This proposition is load-bearing: its contrapositive is used in Corollary 4.14 to produce, for each residue class i ∈ {0,1,...,n-2}, a pair (τ,η) with h_τ = h_η; Lemma 4.15 then uses this to show that h acts 'almost in the same fashion' on any two disjoint cones, and Corollary 4.16 derives the finiteness of local action types. These results are the basis for Lemma 5.1 and hence for Theorems 5.3 and 5.4, the paper's main characterization. Since the manuscript does not prove Proposition 4.13 and [3] is an unpublished preprint, the central claim is not fully verifiable from the manuscript alone. Please provide a proof of Proposition 4.13, or a precise reference to a published version, or at least state the location in [3] where the proof appears and confirm its status.","section":"§4, Proposition 4.13"},{"comment":"The statement 'For n > 3 and 1 ≤ r ≤ n − 1, the group T_{n,r} is infinite' is trivially true, since T_{n,r} is an infinite group. The intended statement, as the surrounding text indicates, is that Out(T_{n,r}) is infinite. Moreover, the proof, which uses the transducers in Figures 2 and 3, requires only n > 2. Please correct the theorem to: 'For n > 2 and 1 ≤ r ≤ n − 1, the group Out(T_{n,r}) is infinite.'","section":"§10, Theorem 10.2"},{"comment":"The theorem states 'The group T_{n,r} for 1 ≤ r < n − 1 has the R∞ property', but the abstract and the introduction claim the R∞ property for all T_{n,r}. The proof in Section 11 appears to cover all 1 ≤ r ≤ n − 1: the orientation-preserving case treats r > 1 explicitly, and the orientation-reversing case treats 1 < r < n. Please correct the range in the theorem statement to '1 ≤ r ≤ n − 1' (with the case n=2, r=1 handled by [7]) and make the abstract and theorem consistent.","section":"§11, Theorem 11.8"}],"minor_comments":[{"comment":"The phrase 'the induced homeomorphisms on Cantor space respects the cyclic ordering' should read 'respect the cyclic ordering' (subject–verb agreement). The same wording appears in Theorem 5.4.","section":"Abstract and Theorem 1.1"},{"comment":"The statement 'For 1 ≤ 1 ≤ n − 1' contains a typo; it should be 'For 1 ≤ i ≤ n − 1'.","section":"Corollary 1.4"},{"comment":"The word 'Proposition' is consistently misspelled as 'Propostion' in the manuscript (e.g., Proposition 4.13, Proposition 7.7). Please correct all occurrences.","section":"Throughout, e.g., §4, Proposition 4.13"},{"comment":"The phrase 'is unto the group of units' should be 'is onto the group of units'.","section":"Theorem 1.10"},{"comment":"Theorem 1.9 states the existence of a homomorphism from X_n to the group of units of Z_{n−1} with kernel X_{n,1} for X = T and G. The proof in §7 gives the construction for O_n (Theorem 7.15) and Remark 7.19 says the results carry over to T O_n without further detail. Since the theorem is stated for both, please add an explicit statement (or one sentence in Remark 7.19) that sig|_{T O_n} is a homomorphism with kernel T O_{n,1}, using Proposition 7.7.","section":"§7, Theorem 1.9 and Remark 7.19"}],"recommendation":"major_revision","confidential_remarks":"The manuscript relies heavily on the BCMNO preprint [3], which is co-authored by the author of this manuscript. This reliance is not circular, but it creates a verifiability issue because the central Proposition 4.13 is not proved in this paper and no published reference is supplied. The editor may wish to ask the author about the current status of [3] and to request that a proof of Proposition 4.13 be included in a revised version or in an appendix. The paper fits the journal's scope and, apart from the local issues identified, is a solid contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Here is what you should know about arXiv:1908.03816. It is a substantive extension of BCMNO's transducer-based automorphism program to the T-family of Higman-Thompson groups. The headline results are real: Aut(T_{n,r}) is characterized as the bi-synchronizing transducers preserving the cyclic order, Out(T_{n,r}) embeds into Out(T_{n,n-1}), all Out(T_{n,r}) for n>2 are infinite and contain a copy of Thompson's group F, and the paper answers BCMNO's gcd question negatively and proves the R-infinity property for T_{n,r}. If correct, this is a major within-field advance, and it goes cleanly beyond Brin-Guzman, whose methods only treated r = n-1.\n\nCredit where due: the paper is careful and detailed, and the adaptation to T is not a routine translation. The transitivity arguments for T are genuinely different from those for G, and the lattice of subgroups Out(X_{n,r}) plus the signature homomorphism are presented surprisingly cleanly. The concrete transducer computations (like the element of TO4 that lies in no TO4,r for r<3) are explicit and checkable.\n\nSoft spots: the main route to finite local actions passes through Proposition 4.13 in the unpublished BCMNO preprint [3]. That proposition is stated without proof, and the stress-test note is right: if it fails, Corollary 4.16 and the finite-transducer characterization collapse. This is a genuine verifiability risk, not an internal inconsistency. The paper needs either a proof of Proposition 4.13 or a public version of [3] that establishes it. There are also two statement-level errors: Theorem 10.2 says \"T_{n,r} is infinite\" where Out(T_{n,r}) was intended, and Theorem 11.8's range \"1 ≤ r < n-1\" does not match the proof, which treats r=1 and then all r>1; the intended range needs a careful fix. The heavy reliance on [3] is understandable given the program, but it makes this paper hostage to that preprint's status.\n\nWho is this for? Specialists in Higman-Thompson groups, automorphism groups of homeomorphism groups, and transducer methods. This is not a paradigm shift, but it is a serious within-field result. I would send it to a competent referee and let them push on the dependence on [3]. My own verdict is conditional: fix the reliance on [3], correct the statement errors, and this paper is publishable.","headline":"Completes the automorphism program for T_{n,r}, but the main proof hinges on an unproven proposition in the unpublished BCMNO preprint.","tokens_in":53842,"tokens_out":3026,"would_cite":true,"duration_ms":32928,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20E36","20F65","20E08"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that the automorphism group of every generalized Thompson group $T_{n,r}$ consists precisely of the finite, bi-synchronizing transducers that preserve the cyclic order on the circle, and derives the structure of the outer…","keywords":["Thompson groups","Higman-Thompson groups","automorphism groups","outer automorphism groups","bi-synchronizing transducers","rational group","R-infinity property","twisted conjugacy"],"falsifier":"Construct a homeomorphism of $C_{n,r}$ that preserves the cyclic-order relation $\\simeq$, lies in the rational group, and has infinitely many distinct local actions $h_\\nu$; if it normalizes $T_{n,r}$, then the paper's finite-local-action corollary fails and Theorem 5.4 cannot hold. The paper's machinery predicts instead that every normalizer element has finitely many local-action types, so this is a direct falsifier.","tokens_in":52902,"feed_emoji":"🔄","tokens_out":13557,"duration_ms":116661,"temperature":0.7,"pith_summary":"The paper sets out to describe, for every $n$ and $r$, the automorphism group of the generalized Thompson group $T_{n,r}$, the circle analogue of the Higman-Thompson groups $G_{n,r}$. Its central theorem identifies $\\mathrm{Aut}(T_{n,r})$ with a concrete subgroup of the rational group: the elements that admit a finite initial transducer which is bi-synchronizing, meaning that both the transducer and its inverse reach a fixed state after reading a long enough prefix, and whose induced homeomorphism of Cantor space preserves the cyclic-order relation $\\simeq$ that collapses Cantor space to the circle. Because every such transducer is also bi-synchronizing in the sense used for $G_{n,r}$, this shows $\\mathrm{Aut}(T_{n,r})$ is a subgroup of $\\mathrm{Aut}(G_{n,r})$. From this description the paper derives the structure of the outer automorphism groups: they embed into a single group $\\mathrm{Out}(T_{n,n-1})$, form a normal lattice indexed by divisibility in $\\mathbb{Z}_{n-1}$, are infinite for $n>2$, and each contains a copy of Thompson's group $F$. The same machinery yields a new infinite family of groups with the $R_\\infty$ property: every automorphism of $T_{n,r}$ has infinitely many twisted conjugacy classes.","feed_headline":"Automorphisms of all generalized Thompson groups are classified","feed_subtitle":"Each one is a finite synchronizing transducer preserving the circle order; outer automorphism groups are infinite and contain F.","key_machinery":"The workhorse is the finite initial transducer over the alphabet $X_n$: a finite automaton with a distinguished initial state that reads one letter from the finite set $\\dot{r}$ and then processes $n$-adic words, emitting $n$-adic words. The load-bearing property is bi-synchronicity: after a fixed-length prefix, the active state of the transducer, and of its inverse, is forced regardless of the starting state. The paper also uses the reduced node distance between incomparable words to encode the cyclic combinatorics of $T_{n,r}$, and the reduced signature of a core transducer, an element of $\\mathbb{Z}_{n-1}$, to control membership in the groups $T O_{n,r}$; the congruence $r\\cdot\\mathrm{sig} \\equiv r \\pmod{n-1}$ organizes the divisibility lattice of outer automorphism groups.","core_discovery":"The central discovery is an exact transducer description of the automorphisms of $T_{n,r}$. Working inside the rational group $R_{n,r}$ of homeomorphisms of the Cantor space $C_{n,r}$ generated by finite initial transducers over the alphabet $X_n$, the paper proves that $\\mathrm{Aut}(T_{n,r})$ is isomorphic to the subgroup $T B_{n,r}$ consisting of those elements of $R_{n,r}$ that can be represented by a finite, initial, bi-synchronizing transducer and whose induced map preserves the equivalence relation $\\simeq$ identifying Cantor space with the circle $S_r$. The proof adapts the corresponding argument for $G_{n,r}$: a normalizer element of $T_{n,r}$ must preserve the tail equivalence relation $\\sim_t$, act on incomparable cones almost in the same fashion, and therefore have only finitely many local actions; these facts force the minimal transducer to be finite and bi-synchronizing, and preservation of $\\simeq$ is shown to be necessary and sufficient. A direct corollary is the strict inclusion $\\mathrm{Aut}(T_{n,r}) < \\mathrm{Aut}(G_{n,r})$.","pith_inferences":["Because both conditions in the characterization are properties of a finite transducer, membership in $\\mathrm{Aut}(T_{n,r})$ and in $\\mathrm{Out}(T_{n,r})$ is decidable for a given element of the rational group; the paper's algorithms for $O_n$ transfer to $T O_n$.","If the signature homomorphism is surjective on $T O_n$, the join of $\\mathrm{Out}(T_{n,r})$ and $\\mathrm{Out}(T_{n,s})$ is exactly $\\mathrm{Out}(T_{n,\\mathrm{lcm}(r,s)})$; surjectivity is proved for $O_n$ whenever divisors of $n$ generate the units of $\\mathbb{Z}_{n-1}$, leaving the $T O_n$ case open.","A natural next step suggested by the methods is an automorphism description for the $F$-analogues of these groups, obtained by running the same local-action and synchronizing argument in the $F$ setting.","The $R_\\infty$ property gives every automorphism of $T_{n,r}$ an infinite Reidemeister number, placing these groups alongside the known examples in the twisted-conjugacy literature."],"forward_implications":["Every automorphism of $T_{n,r}$ is an automorphism of $G_{n,r}$, so $\\mathrm{Aut}(T_{n,r})$ is a subgroup of $\\mathrm{Aut}(G_{n,r})$.","The groups $\\mathrm{Out}(T_{n,r})$ embed in $\\mathrm{Out}(T_{n,n-1})$; divisibility in the cyclic group $\\mathbb{Z}_{n-1}$ gives containments $T O_{n,i} \\leq T O_{n,j}$, with $T O_{n,1}$ contained in every $T O_{n,r}$.","For $n>2$, every $\\mathrm{Out}(T_{n,r})$ is infinite and contains a copy of Thompson's group $F$.","$\\mathrm{Out}(T_{n,r}) = \\mathrm{Out}(T_{n,d})$ whenever $d = \\gcd(n-1,r)$, but the converse isomorphism question has a negative answer (the case $n=7$), so equal gcd is not necessary for isomorphic outer automorphism groups.","Every $T_{n,r}$ has the $R_\\infty$ property: for every automorphism, the twisted conjugacy relation has infinitely many equivalence classes."],"supporting_citations":[{"why":"Supplies the bi-synchronizing transducer characterization of Aut(G_{n,r}) and the local-action lemmas, including the proposition on tail-equivalence preservation, that the proof adapts.","marker":"[3]"},{"why":"Defines the rational group and the minimal-transducer and inverse-construction formalism used throughout.","marker":"[14]"},{"why":"Computes Aut(T_{2,1}) and develops the normalizer approach for Thompson's group T that grounds the method.","marker":"[4]"},{"why":"Establishes Aut(T_{n,n-1}) and the fact that Out(T_{n,n-1}) contains a copy of F, which the paper extends to all r.","marker":"[5]"},{"why":"Proves the R∞ property for Thompson's group F and related groups, providing the base case for the R∞ result.","marker":"[7]"},{"why":"Independently proves the R∞ property for Thompson's group T, the base case the paper extends to T_{n,r}.","marker":"[12]"}],"fun_headline_variants":["Automorphisms of Thompson's T_{n,r} now classified","T_{n,r} automorphisms are bi-synchronizing transducers","Automorphisms of T_{n,r} decoded as finite transducers","Outer automorphism groups of T_{n,r} infinite, contain F"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole transducer description rests on the imported proposition that a homeomorphism of Cantor space whose local actions on $\\tau\\chi$ and $\\eta\\chi$ differ for every suffix $\\chi$ cannot preserve the tail equivalence relation; if that proposition failed, the bi-synchronizing characterization of $\\mathrm{Aut}(T_{n,r})$ would collapse.","fun_headline_variants_meta":{"raw":{"variants":["Automorphisms of Thompson's T_{n,r} now classified","T_{n,r} automorphisms are bi-synchronizing transducers","Automorphisms of T_{n,r} decoded as finite transducers","Outer automorphism groups of T_{n,r} infinite, contain F"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001924,"raw_usage":{"total_tokens":7725,"prompt_tokens":1332,"completion_tokens":6393,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":948,"completion_tokens_details":{"reasoning_tokens":6313}},"tokens_in":948,"tokens_out":6393,"duration_ms":46402,"temperature":1.0,"reasoning_tokens":6313,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:01:23.166413+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct a homeomorphism of $C_{n,r}$ that preserves the cyclic-order relation $\\simeq$, lies in the rational group, and has infinitely many distinct local actions $h_\\nu$; if it normalizes $T_{n,r}$, then the paper's finite-local-action corollary fails and Theorem 5.4 cannot hold. The paper's machinery predicts instead that every normalizer element has finitely many local-action types, so this is a direct falsifier.","supporting_citations":[{"cited_title":"The further chameleon groups of Richard Thompson and Graham Higman: Automorphisms via dynamics for the Higman groups $G_{n,r}$","cited_arxiv_id":"1605.09302","evidence_quote":"Supplies the bi-synchronizing transducer characterization of Aut(G_{n,r}) and the local-action lemmas, including the proposition on tail-equivalence preservation, that the proof adapts."},{"cited_title":"Automata, dynamical systems, and groups","cited_arxiv_id":null,"evidence_quote":"Defines the rational group and the minimal-transducer and inverse-construction formalism used throughout."},{"cited_title":"The chameleon groups of Richard J. Thom pson: automorphisms and dynamics","cited_arxiv_id":null,"evidence_quote":"Computes Aut(T_{2,1}) and develops the normalizer approach for Thompson's group T that grounds the method."},{"cited_title":"Automorphisms of generalized Thompson groups","cited_arxiv_id":null,"evidence_quote":"Establishes Aut(T_{n,n-1}) and the fact that Out(T_{n,n-1}) contains a copy of F, which the paper extends to all r."},{"cited_title":"T he conjugacy problem in extensions of Thompson’s group F","cited_arxiv_id":null,"evidence_quote":"Proves the R∞ property for Thompson's group F and related groups, providing the base case for the R∞ result."},{"cited_title":"Twisted conjugacy in Richard Thompson's group T","cited_arxiv_id":"1309.2875","evidence_quote":"Independently proves the R∞ property for Thompson's group T, the base case the paper extends to T_{n,r}."}],"review_version":1}