TFPT scientific source excerpt Original repository path: experiments/theory-contracts/tfpt-unified-source-derivation-20261029/graded_boundary_functor/GRADED-HS-REFLECTION-AND-TIME.txt Original file SHA-256: 0d9727a9e167c64ef4a8413736c1555b90a7c5da5392eeea3a3455d0bb037986 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 | THE EXISTING GRADED HS SOURCE IMPLEMENTS THE PRODUCT REFLECTION 000002 | 29 October 2026. Analytic continuation, no new model, state fit or test run. 000003 | 000004 | RESULT AND TYPE 000005 | The actual product reflection in EVEN-REFLECTION-POSITIVITY.txt has an exact 000006 | implementation in the already used graded thermal Hilbert-Schmidt source. 000007 | Use the natural CREATION chart of the complex-linear selfdual CAR field. 000008 | It is implemented by J_R=Q_Gamma^* J_HS, Q_Gamma=P_even-i P_odd. Conjugation by this 000009 | antiunitary is ANTILINEAR AND ORDER-PRESERVING. On the whole bounded EVEN 000010 | face algebra its reflection form is a positive HS Gram form. This identifies 000011 | the geometric reflection with a specific graded standard-form partner of 000012 | the GIVEN graph source, including its circle group with the SAME parameter. 000013 | It does not choose that quantization or physical time from P1. 000014 | Euclidean Grassmann word reversal is a separate source/reconstruction map. 000015 | 000016 | 1. Actual h_Phi and its thermal source 000017 | 000018 | Use exactly FINITE-COUPLED.txt: 000019 | g=gamma_F, h=A g-i0 g Phi, 000020 | A=-i0 partial_theta+a(theta), [A,g]=0, 000021 | Phi=Phi^*, {g,Phi}=0. 000022 | For the geometric graph of collar length ell in EVEN-REFLECTION-POSITIVITY.txt, 000023 | fix beta=2ell. Thus the X below is exactly m_ell(h), not a different thermal 000024 | covariance at an independently varied temperature. 000025 | The background is smooth, bounded, classical and normal-independent here. 000026 | No commutation [A,Phi]=0 is needed. In a finite regulator h is the Hermitian 000027 | matrix built from the same actual finite coefficient, including the original 000028 | v252 Yukawa/Majorana block and grading. No replacement coefficient is used. 000029 | 000030 | In the continuum h is selfadjoint on periodic H1. It is a bounded zeroth-order 000031 | perturbation of a first-order elliptic circle operator with finite internal 000032 | rank. Hence its eigenvalues lambda_j have N_|h|(E)=O(E), and 000033 | Tr exp(-beta |h|)0. 000034 | The source Fock presentation is relative to the negative-energy sea of this 000035 | given h. In its excitation basis define the honest positive operator 000036 | H=sum_j |lambda_j| n_j^exc >=0. 000037 | Finite zero modes cause only finite multiplicity. In infinite dimension one 000038 | does NOT subtract two divergent expressions dGamma(h)-Tr(P_- h). 000039 | The positive particle/hole excitation formula defines H on its usual Fock 000040 | domain. The signed charged annihilators satisfy [H,a_j]=-lambda_j a_j: 000041 | positive lambda labels are particle annihilators; negative labels are hole 000042 | creators. Consequently the given free Gibbs source has 000043 | rho=exp(-beta H)/Z_beta, 000044 | Z_beta=product_j(1+exp(-beta |lambda_j|)), 000078 | omega(a(v)a^*(w))=. 000079 | This states the operator-C convention without a hidden transpose. The 000080 | mode formulas below can be read in any fixed orthonormal h eigenbasis. 000081 | The complex-linear geometric CAR field therefore has the natural map 000082 | 000083 | iota[c(T_faces(v0,v1,bar(w0),bar(w1)))] 000084 | =a_R^*(v0)-a_L^*(v1)+a_R(w0)-a_L(w1). (1) 000085 | 000086 | Thus its positive-grade slots are p0(v)=a_R^*(v) and p1(v)=-a_L^*(v). 000087 | The conjugate slots are their actual adjoints. No spectral conjugation 000088 | kappa_h, redefined linear annihilator or extra field half is used. 000089 | For mode labels, ordinary CAR gives 000090 | ==1-C, 000091 | ==C, 000092 | =-==X. 000093 | The other cross block is the adjoint X. Hence the whole geometric covariance is 000094 | P_h=[[1-C,X],[X,C]]. 000095 | The existing GNS intertwiner ARGUMENT, applied to this P_h and rho(h), 000096 | identifies this graph state with its HS vector state on all CAR words. 000097 | It is not a previously proved unitary to the old scalar physical source. 000098 | In infinite dimension it uses bounded smeared 000099 | CAR fields and cyclic closure, not an everywhere bounded inverse of rho. 000100 | 000101 | 3. Exact geometric R, including both odd phases 000102 | 000103 | The existing geometric matrices are 000104 | F=S tensor g, R_geo=face_swap tensor F, 000105 | C0=i0 N, E v=(P_+v,i0 P_-v). 000106 | They obey F E=C0 E(-i0) and F C0=-C0 F. Under the oriented chart 000107 | T_faces(v0,v1,bar(w0),bar(w1)) 000108 | =(Ev0+C0Ew0, Ev1-C0Ew1), 000109 | the full action is 000110 | (v0,v1,bar(w0),bar(w1)) 000111 | ->(-i0 w1,i0 w0,overline(-i0 v1),overline(i0 v0)). 000112 | With the natural creation map (1), this is precisely 000113 | Theta_geo(p0(v))=-i p1(v)^*=+i a_L(v), 000114 | Theta_geo(p1(v))=+i p0(v)^*=+i a_R(v). 000115 | These are antilinear in v on both sides, as required. Equivalently 000116 | Theta_geo(a_R^*(v))=+i a_L(v), 000117 | Theta_geo(a_L^*(v))=-i a_R(v). 000118 | Both adjoint formulas follow by the star operation. The map on the algebra 000119 | is explicitly order-preserving, as in EVEN-REFLECTION-POSITIVITY.txt:59-62. 000120 | 000121 | Put 000122 | Q_Gamma=(1+Gamma)/2-i(1-Gamma)/2 000123 | =exp(-i pi/4) exp(i pi Gamma/4), 000124 | J_R=Q_Gamma^* J_HS. 000125 | Since J_HS is antiunitary and commutes with Gamma, J_R^2=1. Also 000126 | J_R Omega=Omega. For any odd left or right multiplication operator O, 000127 | Q_Gamma^* O Q_Gamma=-i Gamma O. 000128 | Therefore direct operator calculation gives 000129 | J_R a_R^*(v) J_R=-i Gamma R_(a(v))=+i a_L(v), 000130 | J_R[-a_L^*(v)]J_R=+i a_R(v). 000131 | This is the exact geometric formula above, not just a matching cross norm. 000132 | The graph-to-HS GNS unitary for THIS h intertwines J_R with Theta_geo's cyclic 000133 | antiunitary, because it intertwines every bounded field and the cyclic vector. 000134 | Q_Gamma^* is the fixed grading phase forced by those two displayed odd phases. 000135 | It is not the isolated original finite charge J_F, nor is J_R identified 000136 | with a physical OS word reflection by this calculation alone. 000137 | 000138 | 4. Entire even RP form and comparison with the independent Wick proof 000139 | 000140 | For a bounded EVEN source operator A, Q_Gamma commutes with L_A and R_A. 000141 | Thus 000142 | J_R L_A J_R=R_(A^*). 000143 | In particular J_R n_R J_R=1-n_L, not n_L. For any bounded even A_i,A_j, 000144 | 000145 | =Tr(sqrt(rho) A_i^* sqrt(rho) A_j) 000146 | =_HS. 000147 | For i=j this is nonnegative. The Schatten inequality makes the displayed 000148 | vectors HS for bounded A, since rho^(1/4) belongs to Schatten4. This proves 000149 | positivity on the entire even source von Neumann algebra, hence on its 000150 | bounded even CAR subalgebra. The corresponding formula for an even right 000151 | operator gives positivity on the other face as well. 000152 | 000153 | This is the same even reflection and state as in 000154 | EVEN-REFLECTION-POSITIVITY.txt:105-131. There the cross contraction on real 000155 | first-face fields is -i M, and the cross-pairing sign cancels its phase at 000156 | every even Wick length. Here the same statement is the HS Gram identity. 000157 | The real-field -i phase agrees with the two odd transformations in section3. 000158 | Neither proof replaces the actual same-face contractions or the crossing 000159 | blocks. The Wick proof needs no normal traceclass presentation; the present 000160 | HS identification uses precisely the stated normal thermal source class. 000161 | Unbounded currents and point fields retain their existing domain/limit tasks. 000162 | 000163 | 5. The normalized sech factor does not obstruct positive physical heat 000164 | 000165 | For one already quantized fermion let H=E n, E>0. Its positive heat operators 000166 | exp(-sH) exist for every s>=0, at every finite cylinder length. Nevertheless 000167 | Tr(exp(-ell H) a exp(-ell H) a^*) / Tr exp(-2ell H) 000168 | =exp(-ell E)/(1+exp(-2ell E)) 000169 | =1/(2cosh(ell E)). 000170 | Its second derivative at ell=0 is -E^2/2. There is no contradiction: 000171 | varying ell changes both the Gibbs normalization and the thermal state. 000172 | This normalized, parameter-dependent midpoint correlation is not asserted 000173 | to be a fixed positive spectral measure's Laplace transform. The numerator 000174 | already has the ordinary positive-energy heat representation. Consequently 000175 | sech''(0)<0 cannot show that positive heat or a source Hilbert space appears 000176 | only in a half-infinite-cylinder limit. 000177 | 000178 | The same physical heat factors exist for the stated H built from h_Phi. 000179 | For bounded even insertions at 0