TFPT scientific source excerpt Original repository path: experiments/theory-contracts/tfpt-unified-source-derivation-20261029/aps_source_review/C59-DETERMINANT-AND-CROSS-REVIEW.txt Original file SHA-256: 1fe364e6d2f7da9e905c68251d69e40ba183a3600c51548229c5aa7ab3f77d20 Inclusive original lines: 1-200 Line numbers below refer to the original source. This bounded excerpt is source evidence; it is not a new proof/checker run. Conversation metadata and personal execution details are omitted where indicated. 000001 | C59: AP DETERMINANT, CROSS KERNEL AND THE ACTUAL WEAK SOURCE 000002 | Independent targeted review, 30 October 2026. 000003 | Verdict: SCALAR_AP_SPLIT_CORRECT; SAME_OBJECT_AND_ORIGINAL_SELECTION_NOT_PROVED. 000004 | No Atlas or original edit. 000005 | 000006 | 1. Exact conditional determinant formula 000007 | 000008 | Let the normal double be an antiperiodic circle of length 2 ell, let the 000009 | tangential coefficient be constant, and put 000010 | 000011 | T=h^2+mu^2 >=0, E=sqrt(T). 000012 | 000013 | For the scalar second order operator -partial_n^2+T, with all finite internal 000014 | multiplicities retained in Tr, zeta functional calculus gives, up to a 000015 | T-independent normalization, 000016 | 000017 | F_AP(T)=log det_zeta,AP(-partial_n^2+T) 000018 | =Tr[2 ell E+2 log(1+exp(-2 ell E))] 000019 | =Tr[2 log(2 cosh(ell E))]. (1) 000020 | 000021 | For an arbitrary selfadjoint matrix direction Q=dot(T), including 000022 | [T,Q]!=0, trace cyclicity and the Frechet trace identity give 000023 | 000024 | dot(F_AP)=ell Tr[E^-1 tanh(ell E) Q]. (2) 000025 | 000026 | The split in C59 is therefore algebraically correct: 000027 | 000028 | dot(F_bulk)=ell Tr(E^-1 Q), 000029 | 000030 | dot(F_glue) 000031 | =-2 ell Tr[E^-1 exp(-ell E) X_E Q], (3) 000032 | 000033 | X_E=(2 cosh(ell E))^-1. 000034 | 000035 | No commuting approximation is used in (2) or (3). Operator level derivatives 000036 | still require divided differences; only after the trace may they be written 000037 | with f'(T)Q. Equations (2) and the sum of the two split terms have their 000038 | continuous limit at E=0. The bulk and gluing pieces separately contain E^-1 000039 | and require a gap or a common limiting prescription; their singular parts 000040 | cancel in the total AP determinant derivative. 000041 | 000042 | The full Clifford typing remains necessary. The original normal operator is 000043 | 000044 | D=N(partial_n+B), 000045 | 000046 | and the complex linear charged leg is obtained by the already fixed 000047 | multiplication -ND=partial_n+B. Formula (1) is the second order determinant. 000048 | It does not by itself fix the phase and multiplicity of the first order 000049 | determinant or Pfaffian on the complete selfdual source. 000050 | 000051 | 2. The gluing derivative is not the cross operator X 000052 | 000053 | Equation (3) contains X_E as one factor. It is not X_E itself. It is a scalar 000054 | pairing of Q with 000055 | 000056 | K_glue=-2 ell E^-1 exp(-ell E) X_E. (4) 000057 | 000058 | The distinction is already decisive in dimension two. Take 000059 | 000060 | T=diag(a,b), 00. Thus one determinant derivative can 000076 | vanish while the mixed matrix source response is nonzero. The scalar master 000077 | consumer cannot be identified with the whole two face operator response. 000078 | 000079 | There is a second exact boundary. The existing source has 000080 | 000081 | X_h=(2 cosh(ell h))^-1=(2 cosh(ell |h|))^-1. 000082 | 000083 | At mu=0 this is X_E. For the master regulator mu!=0, (3) instead contains 000084 | 000085 | X_mu=(2 cosh(ell sqrt(h^2+mu^2)))^-1, 000086 | 000087 | which is a different kernel unless the source covariance is changed by the 000088 | same regulator. C59's phrase that the master gluing term and the source cross 000089 | block are exactly the same object is therefore valid only after the explicit 000090 | conditions mu=0 and common AP source have been imposed, and even then only X 000091 | as a factor of the scalar weight (4). 000092 | 000093 | 3. The exact positive cross identity is first order 000094 | 000095 | For the antiperiodic first order operator partial_n+h on the circle of length 000096 | 2 ell, its Green branch from one boundary to the opposite boundary is 000097 | 000098 | S_AP(ell,0) 000099 | =exp(-ell h)(1+exp(-2 ell h))^-1 000100 | =(2 cosh(ell h))^-1 000101 | =X_h. (7) 000102 | 000103 | This is an operator identity by functional calculus. It remains true for the 000104 | actual constant full96 weak coefficient h_p(t), without anticommutation or a 000105 | singular pair reduction. Its source derivative is the existing Frechet 000106 | response 000107 | 000108 | dot(X)=D x_(h^2)[dot(h^2)] (8) 000109 | 000110 | in the unregulated mu=0 source chart. Equation (7), rather than the scalar 000111 | determinant derivative, is the direct Green to mixed face map. 000112 | 000113 | The AP determinant remains compatible with it: its gluing derivative is the 000114 | trace pairing of the already supplied direction Q with the weighted cross 000115 | kernel (4). If the original source provides enough independent Q directions, 000116 | the family of all such pairings can recover that weight by trace duality. The 000117 | single actual radial D_Y direction supplies one genuine scalar response, not a 000118 | reconstruction of the full matrix kernel. 000119 | 000120 | 4. AP is source data, not a consequence of matching 000121 | 000122 | The original homogeneous collar fixes the Poisson range 000123 | graph(exp(-ell h)). Together with the positive boundary metric it fixes the 000124 | orthogonal Calderon graph 000125 | 000126 | G=[[A,X_h],[X_h,1-A]]. 000127 | 000128 | This does not select a closed normal spin structure. An invertible periodic 000129 | Green realization has the same Poisson range and the same orthogonal graph, 000130 | while its raw midpoint kernel uses sinh rather than cosh. The statement 000131 | 000132 | demand that master and source agree, therefore choose AP 000133 | 000134 | is a consistency condition added to the two formulas. It is not a derivation 000135 | of AP from either formula. 000136 | 000137 | AP is already genuine data in the named elliptic NS source and period two NS 000138 | cover. Within that scoped source, (7) is original and no target kernel has 000139 | been inserted by hand. The primitive doubled manifold in the original 000140 | theorem at lines 94 to 170 supplies the deck involution and Calderon data, but 000141 | does not state an AP clutch on the closed normal circle. The normal coordinate 000142 | also does not by itself identify its Euclidean translation with reconstructed 000143 | physical time. 000144 | 000145 | The projected master term at paper tfpt-42 section 027 lines 229 to 248 is a 000146 | definition on Ran(P_adm). Applying (1) to it additionally requires: 000147 | 000148 | * an exact homogeneous product collar of normal length 2 ell; 000149 | * the AP normal clutch; 000150 | * reduction of the operator by P_adm; 000151 | * the same charge legs, multiplicities and determinant line; 000152 | * a stated treatment of mu matching the boundary covariance. 000153 | 000154 | Those bindings are not printed in that master definition. 000155 | 000156 | 5. The claimed moving domain does not follow from the reused symbol 000157 | 000158 | The first primitive theorem defines D_ref and D_rel as the even and odd parts 000159 | of a doubled D under tau. The later selfadjointness theorem starts with a 000160 | Dirac anchor D_geo, chooses a modified APS condition from an adapted boundary 000161 | operator A_partial, and proves that a bounded symmetric D_rel perturbation is 000162 | selfadjoint on that same domain. 000163 | 000164 | The repeated symbol D_rel does not prove D_geo=D_ref as a family identity, 000165 | nor that A_partial moves with every constant Phi. C59 also needs an explicit 000166 | tau action to show that the full constant Phi coefficient is even. Therefore 000167 | the original text does not select the moving APS(h) sea and does not exclude 000168 | the already reviewed fixed reference APS source. Fixed and moving domains 000169 | remain distinct source contracts until an original family map binds them. 000170 | 000171 | 6. Strongest actual conclusion and continuation 000172 | 000173 | For the named AP NS source, the full constant weak h_p(t) has an exact mixed 000174 | face operator X by (7), and the existing actual inner fluctuation has the 000175 | noncommuting response (8). This is the strongest operator result. 000176 | 000177 | For an AP product realization of the projected master with mu=0, its existing 000178 | zeta determinant has the exact scalar mixed response (3) along the same 000179 | source direction. It can therefore serve as a real scalar consumer of the 000180 | already given weak source without adding a new kernel. It neither chooses the 000181 | weak coefficient nor reconstructs dot(X). 000182 | 000183 | The next decisive original check is small: identify the actual closed normal 000184 | spin clutch and P_adm reduction used by D_xi in the master determinant. If it 000185 | is the already named AP NS double and the same unregulated h source, (3) is an 000186 | original master response. If it is periodic, differently regulated or only an 000187 | abstract doubled Dirac operator, the equality to the existing source cross 000188 | kernel does not follow. No new state or source family is needed for this 000189 | check. 000190 | 000191 | SOURCES REVIEWED 000192 | 000193 | [OMITTED: non-scientific conversation or execution metadata] 000194 | [OMITTED: non-scientific conversation or execution metadata] 000195 | 000196 | [OMITTED: non-scientific conversation or execution metadata] 000197 | [OMITTED: non-scientific conversation or execution metadata] 000198 | [OMITTED: non-scientific conversation or execution metadata] 000199 | [OMITTED: non-scientific conversation or execution metadata] 000200 | [OMITTED: non-scientific conversation or execution metadata]