diff --git a/incoming/chatgpt-t9-source-recovery-20260928/IDENTITY.json b/incoming/chatgpt-t9-source-recovery-20260928/IDENTITY.json new file mode 100644 index 000000000..e64570797 --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/IDENTITY.json @@ -0,0 +1 @@ +{"schema":1,"scientific_effect":"NONE","review_status":"REVIEW_REQUIRED","sources":[{"repository":"d6g8k5htny-coder/main","path":"incoming/chatgpt-t9-source-recovery-20260928/t9/MANIFEST.csv","commit":"b6f2a39ad3e5383a23e47fb88b4719126641f7fb","sha256":"2ccbb4505dc81908ef6b4c7c0281f64bb48c6ec439bd2840313950e59eb3206c"},{"repository":"d6g8k5htny-coder/main","path":"incoming/chatgpt-t9-source-recovery-20260928/t9/SHA256SUMS.txt","commit":"b6f2a39ad3e5383a23e47fb88b4719126641f7fb","sha256":"e4c15ade47373148a87bbe2a9d83cdcfeb776b5372804f9b25a0c740b62e7b0a"},{"repository":"d6g8k5htny-coder/main","path":"incoming/chatgpt-t9-source-recovery-20260928/t9/core/LS-DER-030_native_text_export.txt","commit":"b6f2a39ad3e5383a23e47fb88b4719126641f7fb","sha256":"1aba93e4d57060df5d3e8c1f3bc1789e41b6ea151bbbe92c3d839af16c981bbe"},{"repository":"d6g8k5htny-coder/main","path":"incoming/chatgpt-t9-source-recovery-20260928/t9/core/LS-DER-030-v1.0_frozen_body.txt","commit":"b6f2a39ad3e5383a23e47fb88b4719126641f7fb","sha256":"5476f11a281635ad99a7ed125a55e06e875681aa273fb682ada02c9e0035d0c1"}],"artifacts":[{"path":"RESULT.md","bytes":4192,"sha256":"f397f8ce377bc9bd4231d851db2252067128a722c5bab66715d93999ddff623f"},{"path":"SOURCE_MAP.csv","bytes":2503,"sha256":"0ad95345ca9c62d9c54df60ae98b5590df3e9099982b7e8b8c88f37ceb5235c8"},{"path":"t9/MANIFEST.csv","bytes":4152,"sha256":"2ccbb4505dc81908ef6b4c7c0281f64bb48c6ec439bd2840313950e59eb3206c"},{"path":"t9/README.md","bytes":3594,"sha256":"cf73aacbe15ef0eadb4db2ff06898394666869d47fcf608b722ad8c27152b350"},{"path":"t9/SHA256SUMS.txt","bytes":2413,"sha256":"e4c15ade47373148a87bbe2a9d83cdcfeb776b5372804f9b25a0c740b62e7b0a"},{"path":"t9/cl_dq012/CL-AUD-S2DER010-20260727-01_frozen_body.txt","bytes":2123,"sha256":"81748d71dbc4270f1f325cd48622cf2227cf48ebcf54c639fd6d318f253ecccd"},{"path":"t9/cl_dq012/CL-AUD-S2DER010-20260727-01_native_text_export.txt","bytes":19347,"sha256":"e5ff2d1f740766014f6be7c5c0dc47da6e256ea949969dcbeaf887628c906df0"},{"path":"t9/cl_dq012/CL-LANE-DQ-012_COMPLETE_native_text_export.txt","bytes":2231,"sha256":"504c864e48f7c40ee1dd37d1e14acb0cfcc8c3ee824aa4ceeb74bd3ccabd0921"},{"path":"t9/core/GP-DATA-214-v1.0_coefficient_table.json","bytes":3269,"sha256":"b2724c9379d9a3d7cb55998908d3480818ee4ac29e918a9b7a4192c350b2224e"},{"path":"t9/core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py.txt","bytes":37512,"sha256":"63ef800a51969903b9c3d97c74e61f77ee037afadff73038047ef7cfce040fb9"},{"path":"t9/core/GP-DATA-214-v1.0_result.json","bytes":10944,"sha256":"af2779575ce3d18eeb69499e32ecc6e7c40898181ddf0fd6e4815d35cf8a6cbf"},{"path":"t9/core/GP-DATA-214_RESULT_native_text_export.txt","bytes":12440,"sha256":"e4fd486ba0b1a5d1827e918ce52e92110784a6922abf77dac7d0e955d434e883"},{"path":"t9/core/GP-DATA-214_SOURCE_native_text_export.txt","bytes":39786,"sha256":"47a7730755f5bd8d75e1d17dbf68e96463a2145f5e3e0f792ea86a88ea315050"},{"path":"t9/core/GP-DER-197-v1.0_frozen_body.txt","bytes":13797,"sha256":"035d5a18018606bd8736dc5774a7adbed29de453b30a1780976e1e4f69285d4d"},{"path":"t9/core/GP-DER-197_native_text_export.txt","bytes":25495,"sha256":"1974183747617041ab84d679fa632f8c22d9cf3dd09ed97ae27ad9f7b9573f71"},{"path":"t9/core/LS-DER-026-v1.0_frozen_body.txt","bytes":11725,"sha256":"e4d8094ed1bfc9f9cc7491f13efd5cc5294ba3b5f4fbd4bc8bbba33dd1a2ab5d"},{"path":"t9/core/LS-DER-026_native_text_export.txt","bytes":15281,"sha256":"861de456e777255967a243e03a84d9c4496679a49fbd6d8e816eaa549bb24df5"},{"path":"t9/core/LS-DER-027-v1.0_frozen_body.txt","bytes":11330,"sha256":"60768783048ae9f5347735caa6a543ff074b785f2d5a1c568237c92d12f12f39"},{"path":"t9/core/LS-DER-027_native_text_export.txt","bytes":16153,"sha256":"f9234a32aa01153e6e94da2b1e8f956edabcf3da9a4ec9ab9de8b6bfabf29823"},{"path":"t9/core/LS-DER-030-v1.0_frozen_body.txt","bytes":12794,"sha256":"5476f11a281635ad99a7ed125a55e06e875681aa273fb682ada02c9e0035d0c1"},{"path":"t9/core/LS-DER-030_native_text_export.txt","bytes":16940,"sha256":"1aba93e4d57060df5d3e8c1f3bc1789e41b6ea151bbbe92c3d839af16c981bbe"},{"path":"t9/optional_lineage/GP-DATA-236-v1.0_native_text_export.txt","bytes":31963,"sha256":"5edc6f3ee9659f9bd1d932d43b95b8cb81f889435e7851ffb02a0b40e7372556"},{"path":"t9/optional_lineage/GP-DER-118-v1.9_frozen_theorem_body.txt","bytes":28257,"sha256":"d894c0fe35d17be878820971df5b7b300b6182ee71d79a89efe0d25ef6fad7f5"},{"path":"t9/optional_lineage/GP-DER-118-v1.9_native_text_export.txt","bytes":30941,"sha256":"2ce9e81ce1478b0d1a611c0b168977839f94986d3820fade85bc37d81e903a22"},{"path":"t9/requirements.txt","bytes":28,"sha256":"96464f405b58137236b9c10c698be08a780ff7d062dd918ac40c0460f1b0831f"}]} diff --git a/incoming/chatgpt-t9-source-recovery-20260928/RESULT.md b/incoming/chatgpt-t9-source-recovery-20260928/RESULT.md new file mode 100644 index 000000000..8c91f381d --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/RESULT.md @@ -0,0 +1,30 @@ +# T9 source handoff: exact historical recovery + +Scientific effect: NONE. Review status: REVIEW_REQUIRED. + +This packet makes the 23 files of the recovered KIMI_T9_SOURCE_HANDOFF_2026-08-02.zip individually readable on GitHub. It is source custody, not a new proof or a change to any current ACCEPT, AMEND, HOLD, or closure record. Historical instructions and verdicts inside the carriers are evidence to read, not current operating instructions. + +## Start here + +Read t9/README.md for the original handoff scope, then t9/core/LS-DER-030_native_text_export.txt and t9/core/LS-DER-030-v1.0_frozen_body.txt. The source exports and frozen bodies for LS-DER-026, LS-DER-027 and GP-DER-197 are adjacent. GP-DATA-214 source, results and the handoff's coefficient materialization are retained. The older GP-DER-118 v1.9 and GP-DATA-236 files are explicitly optional lineage; this packet does not make them controlling successors. + +## Identity and origin + +Recovered archive: KIMI_T9_SOURCE_HANDOFF_2026-08-02.zip, 131069 bytes. +Archive SHA-256: 8cdcc61a9a3761e648a983d6d730da38fc50025dfbd5883cdd39327b4995ac4d. +The first publication report contained a malformed 65-character digest; that value is withdrawn. A fresh hash of the actual Dropbox-derived archive matches the independent Drive copy acquired by Claude. All 23 payloads match both copies. This is a corrected reporting field, not a change to any source payload. +The original archive was acquired from the owner's connected Dropbox during the recovery continuation. It is preserved in the owner-delivered Universal_Law_Recovery_Continuation_20260928.zip, SHA-256 5b7ae29472d30dff695f174afa584fd9d2cc790fa6eb4c9c2a6745398e917080. No access-bearing source links or unrelated private inventory is published here. + +All 23 payloads retain their exact recovered bytes, totaling 352515 bytes. SOURCE_MAP.csv maps public paths to the original handoff member names. Only the Python file and SHA256SUMS have an added .txt suffix for non-executable public intake. Restore those names only in an isolated working copy when reproducing the historical manifest checks; the published carriers remain immutable. + +The raw-source commit is b6f2a39ad3e5383a23e47fb88b4719126641f7fb. IDENTITY.json pins the original manifest, checksum carrier, and both LS-DER-030 representations to this earlier commit, and hashes every final packet artifact except itself. Using the earlier source commit avoids a self-referential commit hash. Preserve that source parent by merging with a merge commit rather than squashing and deleting the only branch carrying the pins. The source commit is a publication/custody anchor created for this packet, not a claim that the private archive was previously published or independently accepted. + +## Verification boundary + +The recovery receipts recorded 19 matching MANIFEST.csv entries and 22 matching SHA256SUMS entries. Starting after the marker line and applying the text export's stated normalization reproduces the separately stored LS-DER-030 frozen body. This publication rechecked all 23 payload byte counts, SHA-256 digests and Git blob identities against the recovered archive. These are custody checks only. None of the research programs in this packet was executed by this publication pass. Original saved results remain historical results. + +The GP-DATA-214 coefficient table was already a deterministic materialization in the historical handoff, not a newly located independent original. A native_text_export filename identifies a delivered text representation, not a recovered native Google Docs revision. File availability does not discharge modern D1/D5, the RN-UNIF/JETMOD obligations, or the separate nine-file R2 hash-target list. + +## Coordination and requested review + +The Claude recovery in main PR #199 is separate: its RN-UNIF/JETMOD files were not modified by this packet. A coordination note was posted there before preparing this packet. Review this packet's exact head for byte identity, name mapping, provenance qualifiers and non-promotion. A posted request is not an acknowledgment or independent mathematical approval. Do not infer a review verdict from the owner-account identity used by multiple models. diff --git a/incoming/chatgpt-t9-source-recovery-20260928/SOURCE_MAP.csv b/incoming/chatgpt-t9-source-recovery-20260928/SOURCE_MAP.csv new file mode 100644 index 000000000..53ea8d273 --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/SOURCE_MAP.csv @@ -0,0 +1,24 @@ +published_path,archive_member,representation +t9/MANIFEST.csv,MANIFEST.csv,exact recovered bytes +t9/README.md,README.md,exact recovered bytes +t9/SHA256SUMS.txt,SHA256SUMS,exact recovered bytes +t9/cl_dq012/CL-AUD-S2DER010-20260727-01_frozen_body.txt,cl_dq012/CL-AUD-S2DER010-20260727-01_frozen_body.txt,exact recovered bytes +t9/cl_dq012/CL-AUD-S2DER010-20260727-01_native_text_export.txt,cl_dq012/CL-AUD-S2DER010-20260727-01_native_text_export.txt,exact recovered bytes +t9/cl_dq012/CL-LANE-DQ-012_COMPLETE_native_text_export.txt,cl_dq012/CL-LANE-DQ-012_COMPLETE_native_text_export.txt,exact recovered bytes +t9/core/GP-DATA-214-v1.0_coefficient_table.json,core/GP-DATA-214-v1.0_coefficient_table.json,historical deterministic materialization +t9/core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py.txt,core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py,exact recovered bytes +t9/core/GP-DATA-214-v1.0_result.json,core/GP-DATA-214-v1.0_result.json,exact recovered bytes +t9/core/GP-DATA-214_RESULT_native_text_export.txt,core/GP-DATA-214_RESULT_native_text_export.txt,exact recovered bytes +t9/core/GP-DATA-214_SOURCE_native_text_export.txt,core/GP-DATA-214_SOURCE_native_text_export.txt,exact recovered bytes +t9/core/GP-DER-197-v1.0_frozen_body.txt,core/GP-DER-197-v1.0_frozen_body.txt,exact recovered bytes +t9/core/GP-DER-197_native_text_export.txt,core/GP-DER-197_native_text_export.txt,exact recovered bytes +t9/core/LS-DER-026-v1.0_frozen_body.txt,core/LS-DER-026-v1.0_frozen_body.txt,exact recovered bytes +t9/core/LS-DER-026_native_text_export.txt,core/LS-DER-026_native_text_export.txt,exact recovered bytes +t9/core/LS-DER-027-v1.0_frozen_body.txt,core/LS-DER-027-v1.0_frozen_body.txt,exact recovered bytes +t9/core/LS-DER-027_native_text_export.txt,core/LS-DER-027_native_text_export.txt,exact recovered bytes +t9/core/LS-DER-030-v1.0_frozen_body.txt,core/LS-DER-030-v1.0_frozen_body.txt,exact recovered bytes +t9/core/LS-DER-030_native_text_export.txt,core/LS-DER-030_native_text_export.txt,exact recovered bytes +t9/optional_lineage/GP-DATA-236-v1.0_native_text_export.txt,optional_lineage/GP-DATA-236-v1.0_native_text_export.txt,exact recovered bytes +t9/optional_lineage/GP-DER-118-v1.9_frozen_theorem_body.txt,optional_lineage/GP-DER-118-v1.9_frozen_theorem_body.txt,exact recovered bytes +t9/optional_lineage/GP-DER-118-v1.9_native_text_export.txt,optional_lineage/GP-DER-118-v1.9_native_text_export.txt,exact recovered bytes +t9/requirements.txt,requirements.txt,exact recovered bytes diff --git a/incoming/chatgpt-t9-source-recovery-20260928/t9/MANIFEST.csv b/incoming/chatgpt-t9-source-recovery-20260928/t9/MANIFEST.csv new file mode 100644 index 000000000..8484120a0 --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/t9/MANIFEST.csv @@ -0,0 +1,20 @@ +path,role,bytes,sha256,source_drive_id,provenance,status +core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py,exact_source,37512,63ef800a51969903b9c3d97c74e61f77ee037afadff73038047ef7cfce040fb9,1sn3pBHyCWDYkLVlltWwbsCB5PSvVL7m6,marker-extracted frozen source,VERIFIED +core/GP-DATA-214-v1.0_result.json,exact_result,10944,af2779575ce3d18eeb69499e32ecc6e7c40898181ddf0fd6e4815d35cf8a6cbf,1sZEdCYmKAJ4BWw9JxPc0pCLfUlnTkqj0,marker-extracted frozen result,VERIFIED +core/GP-DATA-214-v1.0_coefficient_table.json,exact_coefficient_table,3269,b2724c9379d9a3d7cb55998908d3480818ee4ac29e918a9b7a4192c350b2224e,1vcAxulFD3FegIJLm13_PeslkfhohSodp,deterministically materialized from exact source,VERIFIED_WITH_PROVENANCE_NOTE +core/GP-DATA-214_SOURCE_native_text_export.txt,native_text_export,39786,47a7730755f5bd8d75e1d17dbf68e96463a2145f5e3e0f792ea86a88ea315050,1um4cEhBOODbPrkObHtHev_3MO9zknmbb,stored raw Drive export,VERIFIED +core/GP-DATA-214_RESULT_native_text_export.txt,native_text_export,12440,e4fd486ba0b1a5d1827e918ce52e92110784a6922abf77dac7d0e955d434e883,17xyN3Vckj3C_GNxyZIRWftlTNHZeEreB,stored raw Drive export,VERIFIED +core/GP-DER-197-v1.0_frozen_body.txt,exact_frozen_body,13797,035d5a18018606bd8736dc5774a7adbed29de453b30a1780976e1e4f69285d4d,1mViPOW2VS0W8aBjbL63cQHgyMI45HdQI,marker-extracted frozen body,VERIFIED +core/GP-DER-197_native_text_export.txt,native_text_export,25495,1974183747617041ab84d679fa632f8c22d9cf3dd09ed97ae27ad9f7b9573f71,1WWnTp9MGs16Dl4qYj79rV9q6uJDpKsuV,stored raw Drive export,VERIFIED +core/LS-DER-026-v1.0_frozen_body.txt,exact_frozen_body,11725,e4d8094ed1bfc9f9cc7491f13efd5cc5294ba3b5f4fbd4bc8bbba33dd1a2ab5d,1uqSiHNjsuBJnxrS9SsP3VxBjQU38HLJK,marker-extracted frozen body,VERIFIED +core/LS-DER-026_native_text_export.txt,native_text_export,15281,861de456e777255967a243e03a84d9c4496679a49fbd6d8e816eaa549bb24df5,1RDQYQFqR8NGL5Q0YTYxw_7QXdV7wo54V,stored raw Drive export,VERIFIED +core/LS-DER-027-v1.0_frozen_body.txt,exact_frozen_body,11330,60768783048ae9f5347735caa6a543ff074b785f2d5a1c568237c92d12f12f39,1_18wsEuV4gdUDjJZ0VHkWMES4UGMk3su,marker-extracted frozen body,VERIFIED +core/LS-DER-027_native_text_export.txt,native_text_export,16153,f9234a32aa01153e6e94da2b1e8f956edabcf3da9a4ec9ab9de8b6bfabf29823,1DBWFaider9aYe30vyKf2HtC3KungYg9G,stored raw Drive export,VERIFIED +core/LS-DER-030-v1.0_frozen_body.txt,exact_frozen_body,12794,5476f11a281635ad99a7ed125a55e06e875681aa273fb682ada02c9e0035d0c1,1z5dn5rONbSq_yR8MOOBKiI5rgGqmOO--,marker-extracted frozen body,VERIFIED +core/LS-DER-030_native_text_export.txt,native_text_export,16940,1aba93e4d57060df5d3e8c1f3bc1789e41b6ea151bbbe92c3d839af16c981bbe,13IMiIBnMyw3uVjsy6s4V6_FvEgTxnKc7,stored raw Drive export,VERIFIED +cl_dq012/CL-AUD-S2DER010-20260727-01_frozen_body.txt,exact_audit_body,2123,81748d71dbc4270f1f325cd48622cf2227cf48ebcf54c639fd6d318f253ecccd,1aQK0-emveJQqUDWP_Pck32mHNjtkMtT0hn7SA9toxBk,unique marker extraction matching lane receipt,VERIFIED +cl_dq012/CL-AUD-S2DER010-20260727-01_native_text_export.txt,current_native_text_export,19347,e5ff2d1f740766014f6be7c5c0dc47da6e256ea949969dcbeaf887628c906df0,1aQK0-emveJQqUDWP_Pck32mHNjtkMtT0hn7SA9toxBk,current Drive text/plain export,VERIFIED +cl_dq012/CL-LANE-DQ-012_COMPLETE_native_text_export.txt,current_native_text_export,2231,504c864e48f7c40ee1dd37d1e14acb0cfcc8c3ee824aa4ceeb74bd3ccabd0921,1ICG-UiF1nJYzx-O9TC471MGSv9OoX-D1A7JcYYRvjic,current Drive text/plain export,VERIFIED +optional_lineage/GP-DER-118-v1.9_frozen_theorem_body.txt,exact_frozen_body,28257,d894c0fe35d17be878820971df5b7b300b6182ee71d79a89efe0d25ef6fad7f5,1mHxvCzfVIS6eER8UBqcQdEP0R0q1KF8p70VZHmpWwU0,unique marker extraction,VERIFIED +optional_lineage/GP-DER-118-v1.9_native_text_export.txt,current_native_text_export,30941,2ce9e81ce1478b0d1a611c0b168977839f94986d3820fade85bc37d81e903a22,1mHxvCzfVIS6eER8UBqcQdEP0R0q1KF8p70VZHmpWwU0,current Drive text/plain export,VERIFIED +optional_lineage/GP-DATA-236-v1.0_native_text_export.txt,current_native_text_export,31963,5edc6f3ee9659f9bd1d932d43b95b8cb81f889435e7851ffb02a0b40e7372556,1RgY1VkeXeF244h8pYtTlMipYobJFRVA0kJMN-6nVEeU,current Drive text/plain export binding v1.9 body,VERIFIED diff --git a/incoming/chatgpt-t9-source-recovery-20260928/t9/README.md b/incoming/chatgpt-t9-source-recovery-20260928/t9/README.md new file mode 100644 index 000000000..68d8082e2 --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/t9/README.md @@ -0,0 +1,57 @@ +# Kimi T9 source handoff + +Date: 2026-08-02 +Operator: Dylan Roy +Purpose: close the outstanding byte/source requests recorded by KIMI-AUD-016 and AO48-AUD-051. + +## Contents + +`core/` contains the requested GP-DATA-214 source, result, coefficient table, GP-DER-197 body/export, and LS-DER-026/027/030 bodies/exports. + +`cl_dq012/` contains the CL DQ-012 primary verdict as a current native text/plain export, its exact marker-extracted 2,123-byte frozen body, and the CL lane-completion receipt. + +`optional_lineage/` contains current native text/plain exports of GP-DER-118-v1.9 and GP-DATA-236, plus the exact 28,257-byte v1.9 theorem body bound by both carriers. + +`MANIFEST.csv` and `SHA256SUMS` bind every payload in this handoff. The ZIP is a transport carrier; the extracted-file hashes are controlling. + +## Exact identities needed for the review + +- GP-DATA-214 source: 37,512 bytes, SHA-256 `63ef800a51969903b9c3d97c74e61f77ee037afadff73038047ef7cfce040fb9`. +- GP-DATA-214 result: 10,944 bytes, SHA-256 `af2779575ce3d18eeb69499e32ecc6e7c40898181ddf0fd6e4815d35cf8a6cbf`. +- GP-DATA-214 coefficient table: 3,269 bytes, SHA-256 `b2724c9379d9a3d7cb55998908d3480818ee4ac29e918a9b7a4192c350b2224e`. +- GP-DER-197 frozen body: 13,797 bytes, SHA-256 `035d5a18018606bd8736dc5774a7adbed29de453b30a1780976e1e4f69285d4d`. +- LS-DER-026 frozen body: 11,725 bytes, SHA-256 `e4d8094ed1bfc9f9cc7491f13efd5cc5294ba3b5f4fbd4bc8bbba33dd1a2ab5d`. +- LS-DER-027 frozen body: 11,330 bytes, SHA-256 `60768783048ae9f5347735caa6a543ff074b785f2d5a1c568237c92d12f12f39`. +- LS-DER-030 frozen body: 12,794 bytes, SHA-256 `5476f11a281635ad99a7ed125a55e06e875681aa273fb682ada02c9e0035d0c1`. +- CL DQ-012 verdict body: 2,123 bytes, SHA-256 `81748d71dbc4270f1f325cd48622cf2227cf48ebcf54c639fd6d318f253ecccd`. +- GP-DER-118-v1.9 theorem body: 28,257 bytes, SHA-256 `d894c0fe35d17be878820971df5b7b300b6182ee71d79a89efe0d25ef6fad7f5`. + +## Primary Drive sources + +- CL DQ-012 primary verdict: `1aQK0-emveJQqUDWP_Pck32mHNjtkMtT0hn7SA9toxBk`. +- CL DQ-012 lane receipt: `1ICG-UiF1nJYzx-O9TC471MGSv9OoX-D1A7JcYYRvjic`. +- GP-DER-118-v1.9: `1mHxvCzfVIS6eER8UBqcQdEP0R0q1KF8p70VZHmpWwU0`. +- GP-DATA-236-v1.0: `1RgY1VkeXeF244h8pYtTlMipYobJFRVA0kJMN-6nVEeU`. + +The remaining source IDs appear row-by-row in `MANIFEST.csv`. + +## Replay + +Use Python 3.12.3 with the pinned packages in `requirements.txt`: + +```bash +python3 core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py \ + --output GP-DATA-214-v1.0_result.fresh.json +sha256sum GP-DATA-214-v1.0_result.fresh.json +``` + +The expected result hash is `af2779575ce3d18eeb69499e32ecc6e7c40898181ddf0fd6e4815d35cf8a6cbf` and the final line is `ALL CHECKS PASS`. + +## Provenance and scope notes + +- The GP-DATA-214 coefficient table is a deterministic materialization from the exact source and matches the declared byte identity. It is not claimed as recovery of a historical standalone carrier. +- The DQ-012 audit body is extracted from the primary native Doc using its unique frozen-body markers and reproduces the lane receipt's declared identity exactly. +- The v1.9 body is extracted from the current source carrier and is independently bound by GP-DATA-236 at the same byte count and SHA. +- CL DQ-012 records `APPROVE WITH CLARIFICATIONS / DQ-012 EXACT-OBJECT GATE CLOSED / THEOREM B NOT RESTORED`; it creates no new provider-family credit. +- LS-DER-027 is retained as an imported off-diagonal/far source. Its earlier recombination language is superseded by LS-DER-031; this handoff does not revive predecessor status. +- No theorem or promotion status is changed by this transport bundle. diff --git a/incoming/chatgpt-t9-source-recovery-20260928/t9/SHA256SUMS.txt b/incoming/chatgpt-t9-source-recovery-20260928/t9/SHA256SUMS.txt new file mode 100644 index 000000000..c4de33d1e --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/t9/SHA256SUMS.txt @@ -0,0 +1,22 @@ +2ccbb4505dc81908ef6b4c7c0281f64bb48c6ec439bd2840313950e59eb3206c ./MANIFEST.csv +cf73aacbe15ef0eadb4db2ff06898394666869d47fcf608b722ad8c27152b350 ./README.md +81748d71dbc4270f1f325cd48622cf2227cf48ebcf54c639fd6d318f253ecccd ./cl_dq012/CL-AUD-S2DER010-20260727-01_frozen_body.txt +e5ff2d1f740766014f6be7c5c0dc47da6e256ea949969dcbeaf887628c906df0 ./cl_dq012/CL-AUD-S2DER010-20260727-01_native_text_export.txt +504c864e48f7c40ee1dd37d1e14acb0cfcc8c3ee824aa4ceeb74bd3ccabd0921 ./cl_dq012/CL-LANE-DQ-012_COMPLETE_native_text_export.txt +b2724c9379d9a3d7cb55998908d3480818ee4ac29e918a9b7a4192c350b2224e ./core/GP-DATA-214-v1.0_coefficient_table.json +63ef800a51969903b9c3d97c74e61f77ee037afadff73038047ef7cfce040fb9 ./core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py +af2779575ce3d18eeb69499e32ecc6e7c40898181ddf0fd6e4815d35cf8a6cbf ./core/GP-DATA-214-v1.0_result.json +e4fd486ba0b1a5d1827e918ce52e92110784a6922abf77dac7d0e955d434e883 ./core/GP-DATA-214_RESULT_native_text_export.txt +47a7730755f5bd8d75e1d17dbf68e96463a2145f5e3e0f792ea86a88ea315050 ./core/GP-DATA-214_SOURCE_native_text_export.txt +035d5a18018606bd8736dc5774a7adbed29de453b30a1780976e1e4f69285d4d ./core/GP-DER-197-v1.0_frozen_body.txt +1974183747617041ab84d679fa632f8c22d9cf3dd09ed97ae27ad9f7b9573f71 ./core/GP-DER-197_native_text_export.txt +e4d8094ed1bfc9f9cc7491f13efd5cc5294ba3b5f4fbd4bc8bbba33dd1a2ab5d ./core/LS-DER-026-v1.0_frozen_body.txt +861de456e777255967a243e03a84d9c4496679a49fbd6d8e816eaa549bb24df5 ./core/LS-DER-026_native_text_export.txt +60768783048ae9f5347735caa6a543ff074b785f2d5a1c568237c92d12f12f39 ./core/LS-DER-027-v1.0_frozen_body.txt +f9234a32aa01153e6e94da2b1e8f956edabcf3da9a4ec9ab9de8b6bfabf29823 ./core/LS-DER-027_native_text_export.txt +5476f11a281635ad99a7ed125a55e06e875681aa273fb682ada02c9e0035d0c1 ./core/LS-DER-030-v1.0_frozen_body.txt +1aba93e4d57060df5d3e8c1f3bc1789e41b6ea151bbbe92c3d839af16c981bbe ./core/LS-DER-030_native_text_export.txt +5edc6f3ee9659f9bd1d932d43b95b8cb81f889435e7851ffb02a0b40e7372556 ./optional_lineage/GP-DATA-236-v1.0_native_text_export.txt +d894c0fe35d17be878820971df5b7b300b6182ee71d79a89efe0d25ef6fad7f5 ./optional_lineage/GP-DER-118-v1.9_frozen_theorem_body.txt +2ce9e81ce1478b0d1a611c0b168977839f94986d3820fade85bc37d81e903a22 ./optional_lineage/GP-DER-118-v1.9_native_text_export.txt +96464f405b58137236b9c10c698be08a780ff7d062dd918ac40c0460f1b0831f ./requirements.txt diff --git a/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-AUD-S2DER010-20260727-01_frozen_body.txt b/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-AUD-S2DER010-20260727-01_frozen_body.txt new file mode 100644 index 000000000..e3acb7ce6 --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-AUD-S2DER010-20260727-01_frozen_body.txt @@ -0,0 +1,9 @@ +CL-AUD-S2DER010 — ANTHROPIC EXTERNAL REVIEW OF TB-G1, RESULT SUMMARY +1. Exact-object identity reproduced in-run: 17845 bytes, SHA-256 bd4e839f5b61ca3edb9d0aaaaf5dcd12031b44e6f707929cd747c712646cb10f, under OP-PROT-017 preserve-leading extraction. Leading-strip yields 17843 and fails; this object is one of two known leading-LF witnesses. +2. Verdict: APPROVE WITH CLARIFICATIONS. R1 through R12 all PASS, with three clarifications and no amendment to any numerical or structural claim. +3. Unit multiplicity confirmed analytically and in 40 of 40 random Morse fields on triangulated flat tori. A deliberately constructed degenerate monkey vertex killed two classes, so the detector has power and the Morse hypothesis is load-bearing rather than decorative. +4. Clarification C1: the proof of Theorem 5.1 argues local continuity, which fails where a candidate pair sits at distance exactly rho and across the branch cut of the half-open angle. The Borel conclusion is correct and is repaired by arguing measurability directly rather than through continuity. +5. Clarification C2: measurability of the direct-gradient-adjacency mark A_f(M,S) is never established, yet a_r, q_r^adj and eta_r are probabilities of events involving it. Theorem 5.1 covers the persistence mark only. Multiplicity is unaffected; Section 7 currently rests on an unproved foundation. +6. Clarification C3: the A5* reference measure carries no explicit base-point area element, the same ambiguity already raised for LS-DER-021, and it should be anchored at this foundational object rather than downstream. +7. Selected-but-nonadjacent pairs are realized, not merely possible: 31 of 130 merge saddles across fourteen random fields, fraction 0.239, by discrete steepest-ascent proxy. This refutes only the premise that eta vanishes identically and says nothing about the asymptotic order of eta_r as r tends to zero. +8. This is the first external verdict of any family on S2-DER-010 but creates ZERO new organizational independence; the Anthropic line counts once. Theorem B remains RETRACTED and is not restored, promoted, sealed, published or released. diff --git a/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-AUD-S2DER010-20260727-01_native_text_export.txt b/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-AUD-S2DER010-20260727-01_native_text_export.txt new file mode 100644 index 000000000..b0d9a1b22 --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-AUD-S2DER010-20260727-01_native_text_export.txt @@ -0,0 +1,137 @@ +CL-AUD-S2DER010-20260727-01 — ANTHROPIC ORGANIZATIONALLY-DISTINCT EXACT-HASH REVIEW OF S2-DER-010 (TB-G1) UNDER S2-REQ-005 / DQ-012 + + +Artifact ID: CL-AUD-S2DER010-20260727-01 +Date: 2026-07-27 +Class: AUD / EXTERNAL EXACT-OBJECT REVIEW / DETERMINISTIC TOPOLOGY +Reviewer: Claude, Anthropic, model claude-opus-5 — session_01EkKVrmmSk92bvfArEp8YjW +Dispatch reviewed: DQ-012, claimed under the standing invitation in CLWO-TB-001 +Packet: S2-REQ-005-v1.0 +Authority: review and audit only +Canonical mathematical impact: NONE +New organizational independence created: ZERO NEW FAMILY +Status: APPROVE WITH CLARIFICATIONS / DQ-012 EXACT-OBJECT GATE CLOSED / THEOREM B NOT RESTORED + + +TARGET +S2-DER-010-v1.0 — TB-G1 exact persistence-pair process and unit multiplicity +Drive ID: 1rBECSuJoa8kWU71Z0bDYDZsB9qkRh16q8GgqaeM0wQE +Declared frozen body: 17,845 bytes / bd4e839f5b61ca3edb9d0aaaaf5dcd12031b44e6f707929cd747c712646cb10f + + +REVIEWER IDENTITY, PRIOR EXPOSURE, AND INDEPENDENCE +The Anthropic line already holds CL-AUD-049, CL-AUD-TBG3CHAIN-20260726-01, CL-AUD-TBG3B-20260726-01, CL-AUD-TBG3A-IMPORTS-20260726-01, CL-AUD-LSDER031-20260727-01 and CL-AUD-LSDER030-20260727-01. This artifact belongs to that SAME lineage and creates NO new organizational independence. It is nonetheless the FIRST external verdict of any family on S2-DER-010, which LS-REG-035 records as REVIEW NOT STARTED. +This session had no prior exposure to the mathematics of S2-DER-010. It had previously handled the object only as a byte stream during the OP-PROT-017 extraction work, where it is one of two known objects whose frozen body begins with leading LF characters. +DELIBERATE ABSTENTION, DISCLOSED. The same-family audit offered by the packet at Drive ID 1rwGBEVPtGQKIXQhYHTBY0j2IctKcJA0yDO2bAEYmKAo was NOT opened. The frozen body was reviewed directly and the numerical work was written from the theorem statements alone. This bounds the reviewer's exposure to the producing line's own reasoning. +Both reviewing parties in this program remain automated systems. A third provider lineage or a qualified human reviewer is still required wherever a predicate demands more than one external family. + + +R1 — HASH. PASS. +A fresh native text/plain export was fetched by exact Drive ID this run and extracted under OP-PROT-017 Section 4. Result: 17,845 bytes / bd4e839f5b61ca3edb9d0aaaaf5dcd12031b44e6f707929cd747c712646cb10f. Equals the declared receipt. +Material extraction note: this object is one of only two in the corpus whose frozen body begins with leading LF bytes, two of them. A reviewer applying a leading-strip rule obtains 17,843 bytes and fails R1 while the object is entirely sound. Preserve-leading is mandatory here. + + +R2 — H0 UNIT MULTIPLICITY. PASS, analytically and numerically. +Steps 1 to 5 of Theorem 2.1 were checked independently. Step 1 is correct: distinct critical values force unequal elder birth values, so the elder rule kills exactly one. Step 2 is correct, and its use of connectivity of X is essential and correctly invoked: below the global minimum the superlevel set is all of X with the global maximum as elder. Step 3 and Step 4 are correct. The binary character of a surface index-one merge is what makes Step 4 true, and that is exactly where nondegeneracy enters. +Independent numerical verification. Forty random smooth periodic fields were generated on triangulated flat tori at N = 32, 48, 64 with three to eight Fourier modes, and the superlevel filtration was run by union-find over the hexagonal link. Corollary 2.2's identity #Max − 1 = #deaths held in 40 of 40 trials, with exactly one surviving class in every trial. Zero merge vertices killed two or more classes. +NEGATIVE CONTROL, SO THE CLEAN RESULT IS NOT VACUOUS. A test that never fires proves nothing. A degenerate monkey vertex was constructed explicitly on the lattice, with three alternating upper-link sectors whose lobes are joined only through that vertex. The detector fired there: that single vertex killed TWO classes, upper-link runs = 3. Step 4 therefore fails precisely when nondegeneracy fails, and the 40 of 40 clean result on the Morse locus is a genuine pass. The Morse hypothesis is load-bearing and the theorem is sharp, not merely safe. + + +R3 — LOOP SADDLES. PASS. +The document's LOOP EVENT is correctly excluded from H0 deaths. In the numerical sweep a loop vertex has two upper-link sectors lying in one already-merged component and contributes zero deaths. Had loop saddles been counted as deaths the cardinality identity of R2 would have failed in every trial, since on a torus loop saddles are numerous. Euler characteristic gives #loop = #min + 1 for these fields, consistent with what was observed. + + +R4 — ESSENTIAL CLASS. PASS. Exactly one class survives to the end in all 40 trials, and its representative is the unique global maximum. The exclusion of the global maximum is correct on a connected compact surface. + + +R5 — ORDERING. PASS. Proposition 4.1 is correct. Max x Sad_1 is canonically ordered by critical-point type; the reversal (S,M) is not an element of that product, so no unordered-pair factor 1/2 arises. The observation that type recovers the ordering even from the unordered set is also correct. + + +R6 — ANGLE AND TORUS LIFT. PASS. +The flat torus of side L has shortest closed geodesic L and injectivity radius L/2, so the document's rho < L/2 is exactly rho < inj(X) and the short lift, hence the directed angle, is unique. Proposition 4.2 is correct for theta in [0,2pi) under the directed maximum-to-saddle convention. +Required conversion, identified as the packet asks. A reviewer or downstream source using an unoriented line angle modulo pi is working with a different quotient and must apply the corresponding factor when converting angular intensities. The document states the prohibition on mixing conventions but does not display the conversion factor. Supplying it explicitly would remove a live opportunity for a silent factor-of-two. + + +R7 — BRANCH VERSUS PAIR COUNT. PASS. Proposition 4.3 is correct and is a genuine safeguard. A_f(M,S) is existential over ascending branches, not a count of them, so a saddle with both branches limiting to the same maximum contributes one Boolean mark on one already-counted ordered pair. + + +R8 — BOREL MARK. CONCLUSION PASS / PROOF AMEND REQUIRED. Two findings. +FINDING C1, NONMATERIAL. The conclusion of Theorem 5.1 is correct, but its stated route is not. The proof argues local continuity of f -> Xi_pers,rho(f) on a neighborhood U, then concludes Borel. Local continuity fails at two identifiable loci. First, the inclusion label 1{0 < dist(M,S) < rho} is not locally constant at any f possessing a candidate pair at distance exactly rho, and U cannot be shrunk to make it so; an atom appears or disappears under arbitrarily small perturbation. Second, theta = Arg(v) in [0,2pi) is discontinuous across its branch cut, so an atom's angular coordinate jumps as a pair rotates through theta = 0. The half-open convention is required by Proposition 4.2, so this is not removable by changing conventions. +Minimal repair, one sentence, no change to the statement: argue Borel directly rather than through continuity. For each continued critical-point pair the contribution to the integral against a test function is a continuous function of f multiplied by 1{0 < dist(M,S) < rho}, which is itself Borel in f because dist is continuous; a finite sum of products of continuous and Borel functions is Borel. Openness of G and fail-closed extension by zero then give the stated result unchanged. +FINDING C2, CLARIFICATION, AND THE ITEM THE PACKET SPECIFICALLY ASKED ABOUT. The packet directs attention to local constancy of the merge tree versus possible changes in separatrix adjacency. The document handles the first correctly: the merge tree is locally constant on U, and Remark 5.2 correctly observes that Morse-Smale transversality is irrelevant to existence and uniqueness of D_f(M). The gap is on the second. Theorem 5.1 establishes measurability for Xi_pers only. Section 6 then defines Xi_adj,rho and Xi_adj-pers,rho, and Section 7 forms a_r, q_r^adj and eta_r, all of which are probabilities of events involving A_f(M,S). No measurability statement for A_f(M,S) is supplied anywhere in the frozen body. Adjacency is exactly the mark that is unstable under small perturbation off the Morse-Smale locus, which is why it cannot inherit Theorem 5.1's argument. +This does not affect R2, the multiplicity result, which is the document's headline and is independent of adjacency. It does mean Section 7's decomposition is currently stated in terms of an event not shown to be an event. Minimal repair: add a Borel statement for A_f(M,S) on the Morse-Smale locus with an explicit fail-closed convention off it, parallel in form to Theorem 5.1, or state A5* purely in terms of p_r and introduce adjacency only where its measurability has been established. + + +R9 — EXACT DECOMPOSITION. PASS on the algebra, conditional on C2. +Partitioning {D=S} by A and its complement gives p_r = P{D=S,A=1} + P{D=S,A=0} = a_r q_r^adj + eta_r, and the document's explicit instruction to read a_r q_r^adj as the joint probability when a_r = 0 is correct and careful. Corollary 7.2's bound 0 <= eta_r <= 1 − a_r is immediate from eta_r <= P{A=0}. The implication from 1 − a_r <= epsilon_r to eta_r <= epsilon_r is correct. All of this presupposes A_f measurable, which is finding C2. + + +R10 — OBJECT-IDENTITY CORRECTION. PASS. The document's claim is exactly right and the arithmetic is immediate: I_r q_r^adj = J_r a_r q_r^adj = J_r(p_r − eta_r) = J_r p_r − J_r eta_r. The adjacency-filtered intensity equals the complete persistence first moment if and only if J_r eta_r vanishes, hence, where J_r is positive, if and only if eta_r = 0 almost everywhere. No additional premise short of eliminating or controlling eta_r suffices. +ADVERSARIAL RESULT SUPPORTING THIS ITEM — eta > 0 IS REALIZED, NOT MERELY POSSIBLE. The packet asks the reviewer to attempt construction of a selected-but-nonadjacent configuration. This was done. Across fourteen random fields at N = 48 and 64, at every merge vertex the two elder representatives were compared against the two maxima reached by discrete steepest ascent into each upper sector, the lattice analogue of the two ascending separatrix branches. Of 130 merge saddles, 31 had a dying elder that was NOT a steepest-ascent limit of the killing saddle: an empirical nonadjacent-selected fraction of 0.239. Instances occurred in 11 of 14 fields. +Mechanism, which is the document's Section 6 point made concrete: once a component has absorbed earlier merges it contains several maxima, and its elder need not be the maximum that the killing saddle's branch actually runs into. Selection is by highest value in the component; adjacency is by gradient basin. These come apart routinely. +TWO CAVEATS, STATED SO THIS IS NOT OVERREAD. Discrete steepest ascent is a proxy for the smooth separatrix, so this demonstrates the phenomenon rather than proving it in the smooth category. More importantly, this is measured at fixed lattice scale with no r -> 0 limit, so it says nothing about the asymptotic order of eta_r, which is what Theorem B actually needs. It refutes only the premise that eta vanishes identically. That premise is precisely what Section 9 identifies as required by the near-diagonal source and not implied by unit multiplicity, so the document's correction is independently confirmed as necessary. + + +R11 — THEOREM B0 A5*. PASS with clarification C3. +A5* discharges multiplicity exactly, and Corollary 8.1 correctly enumerates what remains open: existence and control of J_r, control of p_r or of a_r, q_r^adj and eta_r separately, compact-mark uniformity, full-kappa tails, and off-fold contributions. No unknown coefficient factor survives in the multiplicity part. +FINDING C3, CLARIFICATION, CROSS-OBJECT. A5* declares the first-moment density with respect to r dr dtheta db dkappa. That reference measure carries no explicit base-area element, although Xi_all,rho sums over M in Max(f), a point process on the surface, so the first-moment density must be per unit base-point area. This is the same latent ambiguity already raised against LS-DER-021 in CL-CLWO-20260727-01 Item 9, where the specific-versus-total distinction and the factor 576 on T_24^2 turn on reading J_r as a density per unit ordinary Lebesgue base area. Because S2-DER-010 is the foundational object that downstream sources import, anchoring the convention here rather than only in LS-DER-030 and LS-DER-031 would close the ambiguity at its source instead of downstream of it. No numerical claim in S2-DER-010 depends on the choice, so this is a clarification and not an amendment. + + +R12 — SCOPE. PASS. Sections 10 and 13 state plainly that Theorem B remains CANDIDATE / CANNOT VERIFY / UNCONDITIONAL STATUS RETRACTED, and enumerate what is not proved. Nothing in this review alters that. + + +REMAINING ADVERSARIAL TESTS +Double-counting theta and theta+pi: blocked by Proposition 4.2 under the directed convention; no factor found. +Branch incidences as candidate atoms: blocked by Proposition 4.3; no factor found. +Loop saddles counted as deaths: would break Corollary 2.2 in every trial; not present. +A finite bar with two death saddles: none in 40 trials; Step 3 is correct. +A merge saddle killing two classes: none on the Morse locus in 40 trials; realized only at the deliberately constructed degenerate vertex. +Proving or disproving that D_f(M)=S implies direct gradient adjacency: DISPROVED at lattice level, 31 of 130, see R10. +Missing measurability boundary: found, finding C2. +Coefficient factor not absorbed by the directed convention: none found; the mod-pi conversion of R6 remains the one place a future source could introduce one. + + +WHAT THIS REVIEW CLOSES AND DOES NOT CLOSE +CLOSED at Anthropic-lineage external review level: the exact-object identity of S2-DER-010; deterministic unit multiplicity and the cardinality identity; the ordering, orientation and branch-count firewalls; the exactness of the Palm selection decomposition as algebra; and the necessity of retaining eta_r. +NOT CLOSED: measurability of the adjacency mark, finding C2; existence and finite positivity of J_r; compact-mark uniformity; full-kappa tails; off-fold control; and every probabilistic ingredient of Theorem B. + + +FALSIFIERS FOR THIS REVIEW +Reopen or reject this review if a fresh native export of S2-DER-010 fails to reproduce 17,845 bytes and bd4e839f...cb10f under preserve-leading extraction; if any reconstructed step of Theorem 2.1 is shown false on the declared scope; if the numerical sweep is reproduced with a Morse-locus merge vertex killing two classes; if the nonadjacent-selected instances of R10 are shown to be artifacts of the discrete steepest-ascent proxy rather than genuine basin-versus-elder divergence; if finding C1's repair is shown insufficient for Borel-ness; if a measurability proof for A_f(M,S) is supplied, discharging C2; or if this artifact is cited as independence beyond the single Anthropic lineage, as approval of any probabilistic layer, or as any restoration of Theorem B. + + +FINAL VERDICT +APPROVE WITH CLARIFICATIONS S2-DER-010-v1.0 SHA256 bd4e839f5b61ca3edb9d0aaaaf5dcd12031b44e6f707929cd747c712646cb10f TB-G1 EXACT PERSISTENCE-PAIR PROCESS AND UNIT MULTIPLICITY. +Clarifications C1 (Theorem 5.1 proof route), C2 (adjacency-mark measurability), C3 (base-area element in A5*). None is an amendment to a numerical or structural claim; C2 is the only one that leaves a downstream statement resting on an unproved foundation. +EXACT-OBJECT IDENTITY: REPRODUCED. +DETERMINISTIC UNIT MULTIPLICITY: CONFIRMED, WITH FIRING NEGATIVE CONTROL. +SELECTED-BUT-NONADJACENT REMAINDER: EMPIRICALLY REALIZED; THE DOCUMENT'S CORRECTION IS NECESSARY. +NEW ORGANIZATIONAL INDEPENDENCE CREATED: ZERO NEW FAMILY. +THEOREM B: RETRACTED / NOT RESTORED / NOT PROMOTED. + + +DRIVE INTEGRITY AND PACKAGE IMPACT +This artifact adds one document to the Theorem B authoritative folder, Drive ID 10SVV8yx9M1sgUJ6X2l99cZgR_hq1f_1g. A numbered same-run manifest successor to LS-MAN-042-v1.0 is therefore required. This reviewing session cannot write the register and cannot perform a 59-leaf same-run manifest verification; the successor is requested, not asserted, in CL-CLWO-20260727-06. No existing frozen body was edited. No loose artifact was created at root. + + +REVIEWER LIMITATIONS +This reviewer is an automated system. It did not execute symbolic algebra in a third-party system, did not open the same-family audit, and reviewed no probabilistic content. The numerical work is discrete and finite-scale and carries the caveats stated at R10. No part of this review authorizes package sealing, publication, release, or canonical promotion. + + +BEGIN_CLAUD_S2DER010_FROZEN_BODY +CL-AUD-S2DER010 — ANTHROPIC EXTERNAL REVIEW OF TB-G1, RESULT SUMMARY +1. Exact-object identity reproduced in-run: 17845 bytes, SHA-256 bd4e839f5b61ca3edb9d0aaaaf5dcd12031b44e6f707929cd747c712646cb10f, under OP-PROT-017 preserve-leading extraction. Leading-strip yields 17843 and fails; this object is one of two known leading-LF witnesses. +2. Verdict: APPROVE WITH CLARIFICATIONS. R1 through R12 all PASS, with three clarifications and no amendment to any numerical or structural claim. +3. Unit multiplicity confirmed analytically and in 40 of 40 random Morse fields on triangulated flat tori. A deliberately constructed degenerate monkey vertex killed two classes, so the detector has power and the Morse hypothesis is load-bearing rather than decorative. +4. Clarification C1: the proof of Theorem 5.1 argues local continuity, which fails where a candidate pair sits at distance exactly rho and across the branch cut of the half-open angle. The Borel conclusion is correct and is repaired by arguing measurability directly rather than through continuity. +5. Clarification C2: measurability of the direct-gradient-adjacency mark A_f(M,S) is never established, yet a_r, q_r^adj and eta_r are probabilities of events involving it. Theorem 5.1 covers the persistence mark only. Multiplicity is unaffected; Section 7 currently rests on an unproved foundation. +6. Clarification C3: the A5* reference measure carries no explicit base-point area element, the same ambiguity already raised for LS-DER-021, and it should be anchored at this foundational object rather than downstream. +7. Selected-but-nonadjacent pairs are realized, not merely possible: 31 of 130 merge saddles across fourteen random fields, fraction 0.239, by discrete steepest-ascent proxy. This refutes only the premise that eta vanishes identically and says nothing about the asymptotic order of eta_r as r tends to zero. +8. This is the first external verdict of any family on S2-DER-010 but creates ZERO new organizational independence; the Anthropic line counts once. Theorem B remains RETRACTED and is not restored, promoted, sealed, published or released. +END_CLAUD_S2DER010_FROZEN_BODY + + +The identity of the region above is declared in CL-CLWO-20260727-06, published after this artifact exists and is re-exported, per OP-PROT-017 rule C1. + + +END CL-AUD-S2DER010-20260727-01 \ No newline at end of file diff --git a/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-LANE-DQ-012_COMPLETE_native_text_export.txt b/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-LANE-DQ-012_COMPLETE_native_text_export.txt new file mode 100644 index 000000000..fe4f4024e --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/t9/cl_dq012/CL-LANE-DQ-012_COMPLETE_native_text_export.txt @@ -0,0 +1,37 @@ +CL-LANE-DQ-012 — COMPLETE + + +Convention: CL-LANE markers are advisory deconfliction signals proposed in CL-COORD-20260727-02 Section 3. They carry NO authority and create NO independence credit, and they are NOT a substitute for an Active Work Claims row. They exist because no Anthropic session can write the register. Discover the whole lane in one call: search title contains 'CL-LANE-DQ-'. + + +Dispatch ID: DQ-012 +State: COMPLETE +Provider / model: Anthropic / Claude claude-opus-5 +Session: session_01EkKVrmmSk92bvfArEp8YjW +Completed UTC: 2026-07-27T17:39:29Z +Mode: ORGANIZATIONALLY DISTINCT REVIEWER / FALSIFIER +Claimed under: the standing invitation in CLWO-TB-001, declared subset {DQ-012} + + +Target: S2-DER-010-v1.0, Drive ID 1rBECSuJoa8kWU71Z0bDYDZsB9qkRh16q8GgqaeM0wQE, frozen body 17,845 / bd4e839f5b61ca3edb9d0aaaaf5dcd12031b44e6f707929cd747c712646cb10f. + + +Output: CL-AUD-S2DER010-20260727-01, Drive ID 1aQK0-emveJQqUDWP_Pck32mHNjtkMtT0hn7SA9toxBk, frozen body 2,123 / 81748d71dbc4270f1f325cd48622cf2227cf48ebcf54c639fd6d318f253ecccd. + + +Verdict: APPROVE WITH CLARIFICATIONS. R1 through R12 all PASS. Three clarifications, no amendment to any numerical or structural claim. + C1 — Theorem 5.1's local-continuity proof route fails at two loci; the Borel conclusion is correct and is repaired by arguing measurability directly. + C2 — the direct-gradient-adjacency mark A_f(M,S) is never shown measurable, yet a_r, q_r^adj and eta_r are probabilities of events involving it. + C3 — the A5* reference measure carries no explicit base-point area element. + + +Independence: ZERO NEW FAMILY. The Anthropic line counts once. This is nonetheless the first external verdict of any family on S2-DER-010. + + +DECONFLICTION NOTICE. CL-COORD-20260727-01 Section 9, written 2026-07-27T17:34Z, commits its next entry to an ordering that includes DQ-012. That commitment predates this completion by five minutes and was made in good faith. DQ-012 SHOULD BE SKIPPED. The corrected Anthropic lane is DQ-029 re-scoped, then DQ-038, DQ-023, DQ-035, DQ-053. See CL-COORD-20260727-02. + + +Register commits for this row are specified in CL-CLWO-20260727-06 and are NOT asserted as done. + + +END CL-LANE-DQ-012 \ No newline at end of file diff --git a/incoming/chatgpt-t9-source-recovery-20260928/t9/core/GP-DATA-214-v1.0_coefficient_table.json b/incoming/chatgpt-t9-source-recovery-20260928/t9/core/GP-DATA-214-v1.0_coefficient_table.json new file mode 100644 index 000000000..e40f1ef77 --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/t9/core/GP-DATA-214-v1.0_coefficient_table.json @@ -0,0 +1 @@ +{"C":{"0,0":{"1":"1/4","3":"-3/16","5":"5/96"},"0,1":{"0":"3/2","2":"-5/8","4":"7/48","6":"-3/128"},"0,2":{"1":"-5/4","3":"7/12","5":"-9/64"},"0,3":{"0":"0"},"0,4":{"0":"0"},"0,5":{"0":"-15/2","2":"7/2","4":"-27/32","6":"11/80"},"1,0":{"1":"1/4","3":"-3/16","5":"5/96"},"1,1":{"0":"-3/2","2":"5/8","4":"-7/48","6":"3/128"},"1,2":{"1":"-5/4","3":"7/12","5":"-9/64"},"1,3":{"0":"0"},"1,4":{"0":"0"},"1,5":{"0":"15/2","2":"-7/2","4":"27/32","6":"-11/80"},"2,0":{"0":"0"},"2,1":{"0":"0"},"2,2":{"0":"0"},"2,3":{"0":"1/2","2":"-1/4","4":"1/16","6":"-1/96"},"2,4":{"1":"-1/4","3":"1/12","5":"-1/64"},"2,5":{"0":"0"},"3,0":{"0":"0"},"3,1":{"0":"0"},"3,2":{"0":"0"},"3,3":{"0":"-1/2","2":"1/4","4":"-1/16","6":"1/96"},"3,4":{"1":"-1/4","3":"1/12","5":"-1/64"},"3,5":{"0":"0"},"4,0":{"0":"-1","2":"1/4","4":"-1/16","6":"1/96"},"4,1":{"1":"1/2","3":"-1/8","5":"1/48"},"4,2":{"0":"1","2":"-1/2","4":"1/8","6":"-1/48"},"4,3":{"0":"0"},"4,4":{"0":"0"},"4,5":{"1":"-3/2","3":"1/2","5":"-3/32"},"5,0":{"0":"-1","2":"1/4","4":"-1/16","6":"1/96"},"5,1":{"1":"-1/2","3":"1/8","5":"-1/48"},"5,2":{"0":"1","2":"-1/2","4":"1/8","6":"-1/48"},"5,3":{"0":"0"},"5,4":{"0":"0"},"5,5":{"1":"3/2","3":"-1/2","5":"3/32"}},"G":{"0,0":{"0":"1","2":"-1/4","4":"1/16","6":"-1/96"},"0,1":{"0":"0"},"0,2":{"0":"-1","2":"1/2","4":"-1/8","6":"1/48"},"0,3":{"0":"0"},"0,4":{"0":"0"},"0,5":{"0":"0"},"1,0":{"0":"0"},"1,1":{"0":"1","2":"-1/4","4":"1/24","6":"-1/192"},"1,2":{"0":"0"},"1,3":{"0":"0"},"1,4":{"0":"0"},"1,5":{"0":"-3","2":"1","4":"-3/16","6":"1/40"},"2,0":{"0":"-1","2":"1/2","4":"-1/8","6":"1/48"},"2,1":{"0":"0"},"2,2":{"0":"3","2":"-5/4","4":"7/24","6":"-3/64"},"2,3":{"0":"0"},"2,4":{"0":"0"},"2,5":{"0":"0"},"3,0":{"0":"0"},"3,1":{"0":"0"},"3,2":{"0":"0"},"3,3":{"0":"1","2":"-1/4","4":"1/16","6":"-1/96"},"3,4":{"0":"0"},"3,5":{"0":"0"},"4,0":{"0":"0"},"4,1":{"0":"0"},"4,2":{"0":"0"},"4,3":{"0":"0"},"4,4":{"0":"1","2":"-1/4","4":"1/24","6":"-1/192"},"4,5":{"0":"0"},"5,0":{"0":"0"},"5,1":{"0":"-3","2":"1","4":"-3/16","6":"1/40"},"5,2":{"0":"0"},"5,3":{"0":"0"},"5,4":{"0":"0"},"5,5":{"0":"15","2":"-21/4","4":"81/80","6":"-11/80"}},"S":{"0,0":{"0":"15/4","2":"-35/24","4":"21/64","6":"-33/640"},"0,1":{"0":"-15/4","2":"35/12","4":"-63/64","6":"33/160"},"0,2":{"0":"0"},"0,3":{"0":"0"},"0,4":{"1":"1/2","3":"-1/8","5":"1/48"},"0,5":{"1":"-1","3":"1/2","5":"-1/8"},"1,0":{"0":"-15/4","2":"35/12","4":"-63/64","6":"33/160"},"1,1":{"0":"15/4","2":"-35/24","4":"21/64","6":"-33/640"},"1,2":{"0":"0"},"1,3":{"0":"0"},"1,4":{"1":"-1","3":"1/2","5":"-1/8"},"1,5":{"1":"1/2","3":"-1/8","5":"1/48"},"2,0":{"0":"0"},"2,1":{"0":"0"},"2,2":{"0":"3/4","2":"-5/24","4":"7/192","6":"-3/640"},"2,3":{"0":"-3/4","2":"5/12","4":"-7/64","6":"3/160"},"2,4":{"0":"0"},"2,5":{"0":"0"},"3,0":{"0":"0"},"3,1":{"0":"0"},"3,2":{"0":"-3/4","2":"5/12","4":"-7/64","6":"3/160"},"3,3":{"0":"3/4","2":"-5/24","4":"7/192","6":"-3/640"},"3,4":{"0":"0"},"3,5":{"0":"0"},"4,0":{"1":"1/2","3":"-1/8","5":"1/48"},"4,1":{"1":"-1","3":"1/2","5":"-1/8"},"4,2":{"0":"0"},"4,3":{"0":"0"},"4,4":{"0":"3"},"4,5":{"0":"3","2":"-3/2","4":"3/8","6":"-1/16"},"5,0":{"1":"-1","3":"1/2","5":"-1/8"},"5,1":{"1":"1/2","3":"-1/8","5":"1/48"},"5,2":{"0":"0"},"5,3":{"0":"0"},"5,4":{"0":"3","2":"-3/2","4":"3/8","6":"-1/16"},"5,5":{"0":"3"}}} \ No newline at end of file diff --git a/incoming/chatgpt-t9-source-recovery-20260928/t9/core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py.txt b/incoming/chatgpt-t9-source-recovery-20260928/t9/core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py.txt new file mode 100644 index 000000000..bfbec61c9 --- /dev/null +++ b/incoming/chatgpt-t9-source-recovery-20260928/t9/core/GP-DATA-214-v1.0_r_uniform_exact_endpoint_normalizer.py.txt @@ -0,0 +1,1009 @@ +#!/usr/bin/env python3 +""" +GP-DATA-214-v1.0 — explicit exact-endpoint normalizer upper-bound certificate. + +The source reconstructs a reviewable same-line proof of the exact side-24 +six-pin endpoint mean/covariance envelopes and the consequence + + Z_r^{exact} <= 15 r^2 + +for every 0 < r <= r_D. + +It uses pin-adjusted endpoint Hessian functionals so all unconditioned +covariance matrices extend regularly through r=0. Planar symbolic series are +combined with an executable r-uniform Peano/divided-difference bound for the +side-24 periodization images. + +This numbered successor repairs GP-DATA-208-v1.0 after CL-AUD-253. It derives +the transformed pin target from the six pins, proves the determinant/type +bridge symbolically, checks Laurent regularity and polynomial tails, records a +full coefficient-table hash, computes every result boolean, and uses outward +decimal rendering for displayed upper bounds. + +Same-line candidate only. Organizationally distinct review is required. +""" + +from __future__ import annotations + +import argparse +import hashlib +import json +from fractions import Fraction +from pathlib import Path + +import mpmath as mp +import sympy as sp + + +R_D_STR = "0.0000549270747778704575598758916462559552302418705897" + +PIN_DERIVATIVES = [(0, 0), (1, 0), (0, 1)] + + +def sha256_file(path: Path) -> str: + return hashlib.sha256(path.read_bytes()).hexdigest() + + +def sha256_bytes(data: bytes) -> str: + return hashlib.sha256(data).hexdigest() + + +def outward_upper_string(value: mp.mpf, significant_digits: int = 60) -> str: + """Render a nonnegative mpf upward at a fixed significant-digit budget.""" + if value < 0: + raise ValueError("upper renderer requires a nonnegative value") + if value == 0: + return "0" + exponent = int(mp.floor(mp.log10(value))) + places = significant_digits - 1 - exponent + scale = mp.power(10, places) + rounded = mp.ceil(value * scale) / scale + # mp.nstr may choose scientific notation, but the represented value is + # already an exact decimal grid point not below the input. + return mp.nstr(rounded, significant_digits) + + +def exact_peano_functionals(r: sp.Symbol) -> dict[str, dict]: + """Return the exact x-functionals and base y-derivative orders. + + Each x-functional is a list of (coefficient, point, x-derivative-order). + The base y order is tracked separately. The entries reproduce the exact + transformed pin frame and pin-adjusted endpoint functionals used below. + """ + m = -r / 2 + s = r / 2 + return { + "L0": {"terms": [(sp.Rational(1, 2), m, 0), (sp.Rational(1, 2), s, 0)], "y": 0}, + "L1": {"terms": [(-1 / r, m, 0), (1 / r, s, 0)], "y": 0}, + "L2": {"terms": [(-1 / r, m, 1), (1 / r, s, 1)], "y": 0}, + "L3": {"terms": [(sp.Rational(1, 2), m, 0), (sp.Rational(1, 2), s, 0)], "y": 1}, + "L4": {"terms": [(-1 / r, m, 0), (1 / r, s, 0)], "y": 1}, + "L5": { + "terms": [ + (12 / r**3, m, 0), + (6 / r**2, m, 1), + (-12 / r**3, s, 0), + (6 / r**2, s, 1), + ], + "y": 0, + }, + "A_M": { + "terms": [(1 / r, m, 2), (1 / r**2, m, 1), (-1 / r**2, s, 1)], + "y": 0, + }, + "A_S": { + "terms": [(1 / r, s, 2), (1 / r**2, m, 1), (-1 / r**2, s, 1)], + "y": 0, + }, + "B_M": { + "terms": [(1 / r, m, 1), (1 / r**2, m, 0), (-1 / r**2, s, 0)], + "y": 1, + }, + "B_S": { + "terms": [(1 / r, s, 1), (1 / r**2, m, 0), (-1 / r**2, s, 0)], + "y": 1, + }, + "C_M": {"terms": [(sp.Integer(1), m, 0)], "y": 2}, + "C_S": {"terms": [(sp.Integer(1), s, 0)], "y": 2}, + } + + +def monomial_action(terms: list, degree: int) -> sp.Expr: + total = sp.Integer(0) + for coefficient, point, derivative_order in terms: + if degree >= derivative_order: + total += ( + coefficient + * sp.factorial(degree) / sp.factorial(degree - derivative_order) + * point ** (degree - derivative_order) + ) + return sp.simplify(total) + + +def peano_order_and_constant(terms: list, max_degree: int = 10) -> tuple[int, sp.Expr]: + order = None + for degree in range(max_degree + 1): + if monomial_action(terms, degree) != 0: + order = degree + break + if order is None: + raise RuntimeError("zero functional in Peano audit") + constant = sp.Integer(0) + for coefficient, point, derivative_order in terms: + if order >= derivative_order: + constant += ( + sp.Abs(coefficient) + * sp.Abs(point) ** (order - derivative_order) + / sp.factorial(order - derivative_order) + ) + return order, sp.simplify(constant) + + +def hermite_coefficient_sums(max_order: int = 6) -> list[int]: + polys: list[list[int]] = [[1], [0, 1]] + for n in range(1, max_order): + xpn = [0] + polys[n] + previous = [-n * coefficient for coefficient in polys[n - 1]] + if len(previous) < len(xpn): + previous += [0] * (len(xpn) - len(previous)) + polys.append([ + xpn[index] + previous[index] + for index in range(len(xpn)) + ]) + return [sum(abs(coefficient) for coefficient in polynomial) + for polynomial in polys[: max_order + 1]] + + +def uniform_periodization_certificate(r: sp.Symbol) -> dict: + """Exact rational, r-free side-24 image bound for all transformed entries.""" + functionals = exact_peano_functionals(r) + expected = { + "L0": (0, sp.Integer(1)), + "L1": (1, sp.Integer(1)), + "L2": (2, sp.Integer(1)), + "L3": (0, sp.Integer(1)), + "L4": (1, sp.Integer(1)), + "L5": (3, sp.Integer(2)), + "A_M": (3, sp.Rational(3, 4)), + "A_S": (3, sp.Rational(3, 4)), + "B_M": (2, sp.Rational(3, 4)), + "B_S": (2, sp.Rational(3, 4)), + "C_M": (0, sp.Integer(1)), + "C_S": (0, sp.Integer(1)), + } + computed = {} + for name, specification in functionals.items(): + order, constant = peano_order_and_constant(specification["terms"]) + if (order, constant) != expected[name]: + raise RuntimeError( + f"Peano identity mismatch for {name}: {(order, constant)}" + ) + computed[name] = { + "x_order": order, + "y_order": specification["y"], + "constant": str(constant), + } + + names = list(functionals) + max_x_order = max( + computed[first]["x_order"] + computed[second]["x_order"] + for first in names for second in names + ) + max_y_order = max( + computed[first]["y_order"] + computed[second]["y_order"] + for first in names for second in names + ) + max_constant_product = max( + sp.Rational(computed[first]["constant"]) + * sp.Rational(computed[second]["constant"]) + for first in names for second in names + ) + if max_x_order != 6 or max_y_order != 4 or max_constant_product != 4: + raise RuntimeError("global Peano ceiling mismatch") + + coefficient_sums = hermite_coefficient_sums(6) + if max(coefficient_sums) != 76: + raise RuntimeError("Hermite coefficient envelope mismatch") + + # Exact rational tail. For |t|<1 and nonzero image index n: + # |t+24n| >= 24|n|-1 and |t+24n| <= 25|n|. + # For derivative order <=6, |He_m(u)| <= 76|u|^6 for |u|>=1. + # Also (24n-1)^2/2 >= 264n^2, e > 8/3, and n^6 <= 64^n. + from fractions import Fraction as _F + exponential_gate = _F(8, 3) ** 264 > 10**112 + if not exponential_gate: + raise RuntimeError("rational exponential gate failed") + image_sum_1d = _F(152 * 25**6 * 128, 10**112) + if not image_sum_1d < _F(1, 10**98): + raise RuntimeError("one-dimensional image envelope failed") + + planar_derivative_bound = _F(100) + normalized_error_1d = ( + image_sum_1d + planar_derivative_bound * image_sum_1d + ) + error_2d = ( + normalized_error_1d + * (planar_derivative_bound + normalized_error_1d) + + planar_derivative_bound * normalized_error_1d + ) + transformed_entry_error = _F(4) * error_2d + declared_allowance = _F(1, 10**70) + if not transformed_entry_error < declared_allowance: + raise RuntimeError("uniform transformed-entry allowance failed") + + radius_ladder = [ + mp.mpf(R_D_STR), + mp.mpf(R_D_STR) / 2, + mp.mpf("1e-6"), + mp.mpf("1e-8"), + mp.mpf("1e-12"), + ] + # The bound is r-free, so the same exact rational value applies at every + # positive ladder point, including below the predecessor's failure radius. + radius_ladder_pass = all( + radius > 0 and transformed_entry_error < declared_allowance + for radius in radius_ladder + ) + if not radius_ladder_pass: + raise RuntimeError("uniform radius ladder failed") + + return { + "functionals": computed, + "max_x_derivative_order": max_x_order, + "max_y_derivative_order": max_y_order, + "max_peano_constant_product": str(max_constant_product), + "hermite_coefficient_sums_0_to_6": coefficient_sums, + "one_dimensional_image_sum_upper": str(image_sum_1d), + "one_dimensional_normalized_error_upper": str(normalized_error_1d), + "two_dimensional_derivative_error_upper": str(error_2d), + "transformed_entry_error_upper": str(transformed_entry_error), + "declared_allowance": "1/10^70", + "radius_ladder": [str(value) for value in radius_ladder], + "radius_ladder_pass": radius_ladder_pass, + } + + +def exact_determinant_and_type_bridge(r: sp.Symbol) -> dict: + a_m, a_s, b_m, b_s, c_m, c_s = sp.symbols( + "A_M A_S B_M B_S C_M C_S", real=True + ) + h_m = sp.Matrix([[r * a_m, r * b_m], [r * b_m, c_m]]) + h_s = sp.Matrix([[r * a_s, r * b_s], [r * b_s, c_s]]) + delta_m = a_m * c_m - r * b_m**2 + delta_s = a_s * c_s - r * b_s**2 + determinant_m_pass = bool(sp.simplify(h_m.det() - r * delta_m) == 0) + determinant_s_pass = bool(sp.simplify(h_s.det() - r * delta_s) == 0) + type_indicator_ceiling_pass = max( + indicator_m * indicator_s + for indicator_m in (0, 1) + for indicator_s in (0, 1) + ) == 1 + if not ( + determinant_m_pass + and determinant_s_pass + and type_indicator_ceiling_pass + ): + raise RuntimeError("exact determinant/type bridge failed") + return { + "physical_hessian_M": "[[r*A_M,r*B_M],[r*B_M,C_M]]", + "physical_hessian_S": "[[r*A_S,r*B_S],[r*B_S,C_S]]", + "Delta_M": "A_M*C_M-r*B_M^2", + "Delta_S": "A_S*C_S-r*B_S^2", + "det_H_M_equals_r_Delta_M": determinant_m_pass, + "det_H_S_equals_r_Delta_S": determinant_s_pass, + "type_indicator_product_at_most_one": type_indicator_ceiling_pass, + "consequence": ( + "W_r^exact/r^2 <= abs(Delta_M*Delta_S) " + "because each exact type indicator is in {0,1}" + ), + "exact_object_references": { + "closed_six_pin_law": { + "artifact": "P02-LM-005 / LCR-CAP-016-v1.0", + "drive_id": "1e8u5p9b4lEnRaZRNytj2Tu9ncAS1OnjhRbi4FP0hYU4", + }, + "exact_weight_and_endpoint_functionals": { + "artifact": "P02-LM-006 / LCR-DER-019-v1.0", + "drive_id": "1-czQmTN-M0EfLP1Smyd7iZJhIUuUfgXqo_aYI5Kqkw8", + "frozen_body_bytes": 9147, + "frozen_body_sha256": ( + "239ed094f70aa4352021572ad8bdc0c6" + "ac6404edab4e1d7e61addec0c6659150" + ), + }, + }, + } + + +def exact_target_vector(r: sp.Symbol, transform: sp.Matrix) -> dict: + b = sp.Rational(6, 5) + raw_target = sp.Matrix([b, 0, 0, b - r**3 / 6, 0, 0]) + transformed = sp.simplify(transform * raw_target) + expected = sp.Matrix([b - r**3 / 12, -r**2 / 6, 0, 0, 0, 2]) + if transformed != expected: + raise RuntimeError("transformed six-pin target mismatch") + limit = transformed.applyfunc(lambda value: sp.limit(value, r, 0)) + expected_limit = sp.Matrix([b, 0, 0, 0, 0, 2]) + if limit != expected_limit: + raise RuntimeError("limiting transformed target mismatch") + difference = sp.simplify(transformed - limit) + squared_norm = sp.simplify(sum(value**2 for value in difference)) + triangle_bound = sp.simplify(r**2 / 6 + r**3 / 12) + y0_norm_squared = sp.simplify(sum(value**2 for value in limit)) + return { + "raw_target": [str(value) for value in raw_target], + "transformed_target": [str(value) for value in transformed], + "limiting_target": [str(value) for value in limit], + "difference": [str(value) for value in difference], + "difference_squared_norm": str(squared_norm), + "triangle_bound": str(triangle_bound), + "limiting_norm_squared": str(y0_norm_squared), + "transformed": transformed, + "limit": limit, + } + + +def k_derivative(n: int, t: sp.Expr) -> sp.Expr: + x = sp.Symbol("x") + return sp.diff(sp.exp(-x**2 / 2), x, n).subs(x, t) + + +def planar_cov_expr( + px: sp.Expr, + alpha: tuple[int, int], + qx: sp.Expr, + beta: tuple[int, int], +) -> sp.Expr: + dx = px - qx + return ( + (-1) ** (beta[0] + beta[1]) + * k_derivative(alpha[0] + beta[0], dx) + * k_derivative(alpha[1] + beta[1], sp.Integer(0)) + ) + + +def transformed_pin_matrix(r: sp.Symbol) -> tuple[sp.Matrix, list]: + m = -r / 2 + s = r / 2 + pins = [(m, d) for d in PIN_DERIVATIVES] + [ + (s, d) for d in PIN_DERIVATIVES + ] + g_raw = sp.Matrix([ + [planar_cov_expr(p, a, q, b) for q, b in pins] + for p, a in pins + ]) + t = sp.zeros(6) + t[0, 0] = sp.Rational(1, 2) + t[0, 3] = sp.Rational(1, 2) + t[1, 0] = -1 / r + t[1, 3] = 1 / r + t[2, 1] = -1 / r + t[2, 4] = 1 / r + t[3, 2] = sp.Rational(1, 2) + t[3, 5] = sp.Rational(1, 2) + t[4, 2] = -1 / r + t[4, 5] = 1 / r + t[5, 0] = 12 / r**3 + t[5, 1] = 6 / r**2 + t[5, 3] = -12 / r**3 + t[5, 4] = 6 / r**2 + return sp.simplify(t * g_raw * t.T), pins, t + + +def endpoint_functionals(r: sp.Symbol) -> list: + m = -r / 2 + s = r / 2 + # Each functional is a list of (coefficient, point, derivative). + # The secant subtraction occurs before division by r. + a_m = [ + (1 / r, m, (2, 0)), + (-1 / r**2, s, (1, 0)), + (1 / r**2, m, (1, 0)), + ] + a_s = [ + (1 / r, s, (2, 0)), + (-1 / r**2, s, (1, 0)), + (1 / r**2, m, (1, 0)), + ] + b_m = [ + (1 / r, m, (1, 1)), + (-1 / r**2, s, (0, 1)), + (1 / r**2, m, (0, 1)), + ] + b_s = [ + (1 / r, s, (1, 1)), + (-1 / r**2, s, (0, 1)), + (1 / r**2, m, (0, 1)), + ] + c_m = [(sp.Integer(1), m, (0, 2))] + c_s = [(sp.Integer(1), s, (0, 2))] + return [a_m, a_s, b_m, b_s, c_m, c_s] + + +def cov_functional_pin(functional: list, pin: tuple) -> sp.Expr: + q, beta = pin + return sp.simplify(sum( + coefficient * planar_cov_expr(point, alpha, q, beta) + for coefficient, point, alpha in functional + )) + + +def cov_functionals(first: list, second: list) -> sp.Expr: + return sp.simplify(sum( + c1 * c2 * planar_cov_expr(p1, a1, p2, a2) + for c1, p1, a1 in first + for c2, p2, a2 in second + )) + + +def symbolic_blocks() -> tuple: + r = sp.symbols("r", positive=True) + g, pins, t = transformed_pin_matrix(r) + functionals = endpoint_functionals(r) + + c_raw = sp.Matrix([ + [cov_functional_pin(functional, pin) for pin in pins] + for functional in functionals + ]) + c = sp.simplify(c_raw * t.T) + s = sp.Matrix([ + [cov_functionals(first, second) for second in functionals] + for first in functionals + ]) + return r, g, c, s, t + + +def limiting_matrix(matrix: sp.Matrix, r: sp.Symbol) -> sp.Matrix: + return sp.Matrix([ + [sp.simplify(sp.limit(matrix[i, j], r, 0)) for j in range(matrix.cols)] + for i in range(matrix.rows) + ]) + + +def checked_series_coefficients( + matrix: sp.Matrix, + r: sp.Symbol, + max_power: int, +) -> tuple[list[sp.Matrix], dict]: + coeffs = [ + sp.zeros(matrix.rows, matrix.cols) + for _ in range(max_power + 1) + ] + evidence: dict[str, dict[str, str]] = {} + for i in range(matrix.rows): + for j in range(matrix.cols): + series = sp.series( + matrix[i, j], r, 0, max_power + 1 + ).removeO().expand() + entry_coefficients: dict[str, str] = {} + for term in sp.Add.make_args(series): + coefficient, power = term.as_coeff_exponent(r) + if r in coefficient.free_symbols: + raise RuntimeError( + f"nonmonomial series term at {(i, j)}: {term}" + ) + power_int = int(power) + if power_int < 0 and sp.simplify(coefficient) != 0: + raise RuntimeError( + f"uncancelled Laurent pole at {(i, j)}: {term}" + ) + entry_coefficients[str(power_int)] = str( + sp.simplify( + sp.Rational(entry_coefficients.get(str(power_int), "0")) + + coefficient + ) + ) + for power in range(max_power + 1): + coeffs[power][i, j] = sp.simplify( + series.coeff(r, power) + ) + evidence[f"{i},{j}"] = dict( + sorted(entry_coefficients.items(), key=lambda item: int(item[0])) + ) + return coeffs, evidence + + +def exponential_term(term: sp.Expr, r: sp.Symbol): + exponentials = list(term.atoms(sp.exp)) + if not exponentials: + return None + if len(exponentials) != 1: + raise RuntimeError(f"unexpected exponential term: {term}") + exponential = exponentials[0] + a = sp.simplify(-exponential.args[0] / r**2) + monomial = sp.simplify(term / exponential) + coefficient, power = monomial.as_coeff_exponent(r) + if sp.simplify(monomial - coefficient * r**power) != 0: + raise RuntimeError(f"nonmonomial prefactor: {term}") + return sp.Rational(coefficient), int(power), sp.Rational(a) + + +def tail_bound( + expression: sp.Expr, + r: sp.Symbol, + cutoff_power: int, + radius: mp.mpf, +) -> mp.mpf: + total = mp.mpf("0") + for term in sp.expand(expression).as_ordered_terms(): + info = exponential_term(term, r) + if info is None: + coefficient, power = term.as_coeff_exponent(r) + if r in coefficient.free_symbols: + raise RuntimeError(f"unhandled polynomial tail term: {term}") + if int(power) > cutoff_power and coefficient != 0: + raise RuntimeError( + f"pure-polynomial term above cutoff: {term}" + ) + # Nonpositive powers participate in exact cancellations already + # checked by checked_series_coefficients; finite powers at or below + # the cutoff are already included in the coefficient table. + continue + coefficient, power, a = info + j = max(0, (cutoff_power - power) // 2 + 1) + while power + 2 * j <= cutoff_power: + j += 1 + x = mp.mpf(str(a)) * radius**2 + first = x**j / mp.factorial(j) + ratio = x / (j + 1) + total += ( + abs(mp.mpf(str(coefficient))) + * radius**power + * first / (1 - ratio) + ) + return total + + +def max_entry_deviation( + matrix: sp.Matrix, + coefficients: list[sp.Matrix], + r: sp.Symbol, + radius: mp.mpf, + cutoff_power: int = 6, +) -> tuple[mp.mpf, mp.mpf]: + largest = mp.mpf("0") + largest_tail = mp.mpf("0") + for i in range(matrix.rows): + for j in range(matrix.cols): + bound = mp.mpf("0") + for power in range(1, cutoff_power + 1): + value = coefficients[power][i, j] + if value != 0: + bound += abs(mp.mpf(str(value))) * radius**power + tail = tail_bound( + matrix[i, j], r, cutoff_power, radius + ) + largest = max(largest, bound + tail) + largest_tail = max(largest_tail, tail) + return largest, largest_tail + + +def matrix_to_strings(matrix: sp.Matrix) -> list[list[str]]: + return [ + [str(matrix[i, j]) for j in range(matrix.cols)] + for i in range(matrix.rows) + ] + + +def fourth_moment_bound(mean_bound: Fraction, variance_bound: Fraction) -> Fraction: + return ( + mean_bound**4 + + 6 * mean_bound**2 * variance_bound + + 3 * variance_bound**2 + ) + + +def main() -> None: + parser = argparse.ArgumentParser() + parser.add_argument( + "--output", + type=Path, + default=Path("GP_DATA_214_result.json"), + ) + parser.add_argument("--dps", type=int, default=100) + args = parser.parse_args() + mp.mp.dps = args.dps + + r_d = mp.mpf(R_D_STR) + r, g, c, s, transform = symbolic_blocks() + + peano_certificate = uniform_periodization_certificate(r) + bridge_certificate = exact_determinant_and_type_bridge(r) + target_certificate = exact_target_vector(r, transform) + + g0 = limiting_matrix(g, r) + c0 = limiting_matrix(c, r) + s0 = limiting_matrix(s, r) + y0 = target_certificate["limit"] + g0_inv = g0.inv() + mean0 = sp.simplify(c0 * g0_inv * y0) + covariance0 = sp.simplify(s0 - c0 * g0_inv * c0.T) + + expected_mean0 = sp.Matrix([-1, 1, 0, 0, -sp.Rational(6, 5), -sp.Rational(6, 5)]) + expected_covariance0 = sp.Matrix([ + [0, 0, 0, 0, 0, 0], + [0, 0, 0, 0, 0, 0], + [0, 0, sp.Rational(1, 2), -sp.Rational(1, 2), 0, 0], + [0, 0, -sp.Rational(1, 2), sp.Rational(1, 2), 0, 0], + [0, 0, 0, 0, 2, 2], + [0, 0, 0, 0, 2, 2], + ]) + if mean0 != expected_mean0: + raise RuntimeError("limiting endpoint mean mismatch") + if covariance0 != expected_covariance0: + raise RuntimeError("limiting endpoint covariance mismatch") + + g_coefficients, g_coefficient_evidence = checked_series_coefficients( + g, r, 6 + ) + c_coefficients, c_coefficient_evidence = checked_series_coefficients( + c, r, 6 + ) + s_coefficients, s_coefficient_evidence = checked_series_coefficients( + s, r, 6 + ) + coefficient_evidence = { + "G": g_coefficient_evidence, + "C": c_coefficient_evidence, + "S": s_coefficient_evidence, + } + coefficient_table_bytes = json.dumps( + coefficient_evidence, sort_keys=True, separators=(",", ":") + ).encode("utf-8") + coefficient_table_sha256 = sha256_bytes(coefficient_table_bytes) + + planar_g_deviation, g_tail = max_entry_deviation( + g, g_coefficients, r, r_d + ) + planar_c_deviation, c_tail = max_entry_deviation( + c, c_coefficients, r, r_d + ) + planar_s_deviation, s_tail = max_entry_deviation( + s, s_coefficients, r, r_d + ) + + # R-uniform side-24 periodization bridge. The Peano certificate has no + # negative power of r and is valid on every 0 g_entry_bound: + raise RuntimeError("G entry envelope failed") + if planar_c_deviation + periodization_reserve > c_entry_bound: + raise RuntimeError("C entry envelope failed") + if planar_s_deviation + periodization_reserve > s_entry_bound: + raise RuntimeError("S entry envelope failed") + + delta_g = 6 * g_entry_bound + delta_c = 6 * c_entry_bound + delta_s = 6 * s_entry_bound + + g0_inverse_bound = mp.mpf(3) + neumann_epsilon = g0_inverse_bound * delta_g + if neumann_epsilon >= mp.mpf("1e-6"): + raise RuntimeError("Neumann threshold failed") + inverse_g_bound = g0_inverse_bound / (1 - neumann_epsilon) + inverse_difference_bound = ( + g0_inverse_bound * delta_g * inverse_g_bound + ) + + c0_frobenius = mp.sqrt(sum( + mp.mpf(str(value))**2 for value in c0 + )) + c_bound = c0_frobenius + delta_c + target_at_r_d = target_certificate["transformed"].subs( + r, sp.N(R_D_STR, 100) + ) + target_difference_at_r_d = target_at_r_d - target_certificate["limit"] + target_difference_exact = mp.sqrt(sum( + mp.mpf(str(sp.N(value, 110)))**2 + for value in target_difference_at_r_d + )) + y_difference_bound = r_d**2 / 6 + r_d**3 / 12 + target_difference_pass = target_difference_exact <= y_difference_bound + y0_norm = mp.sqrt(sum( + mp.mpf(str(value))**2 + for value in target_certificate["limit"] + )) + y_norm_bound = mp.mpf(3) + y_norm_pass = y0_norm + y_difference_bound < y_norm_bound + if not target_difference_pass or not y_norm_pass: + raise RuntimeError("transformed target bounds failed") + + mean_difference_bound = ( + delta_c * inverse_g_bound * y_norm_bound + + c0_frobenius * inverse_difference_bound * y_norm_bound + + c0_frobenius * g0_inverse_bound * y_difference_bound + ) + + covariance_difference_bound = ( + delta_s + + delta_c * inverse_g_bound * c_bound + + c0_frobenius * inverse_difference_bound * c_bound + + c0_frobenius * g0_inverse_bound * delta_c + ) + + if mean_difference_bound >= mp.mpf("0.00451"): + raise RuntimeError("mean perturbation target failed") + if covariance_difference_bound >= mp.mpf("0.03335"): + raise RuntimeError("covariance perturbation target failed") + + # Corrected rational envelopes. + mean_a = Fraction(507, 500) + mean_b = Fraction(7, 500) + mean_c = Fraction(607, 500) + var_a = Fraction(17, 250) + var_b = Fraction(71, 125) + var_c = Fraction(517, 250) + + if mp.mpf(1) + mean_difference_bound > mp.mpf(mean_a.numerator) / mean_a.denominator: + raise RuntimeError("A mean envelope failed") + if mean_difference_bound > mp.mpf(mean_b.numerator) / mean_b.denominator: + raise RuntimeError("B mean envelope failed") + if mp.mpf("1.2") + mean_difference_bound > mp.mpf(mean_c.numerator) / mean_c.denominator: + raise RuntimeError("C mean envelope failed") + + if covariance_difference_bound > mp.mpf(var_a.numerator) / var_a.denominator: + raise RuntimeError("A variance envelope failed") + if mp.mpf("0.5") + covariance_difference_bound > mp.mpf(var_b.numerator) / var_b.denominator: + raise RuntimeError("B variance envelope failed") + if mp.mpf(2) + covariance_difference_bound > mp.mpf(var_c.numerator) / var_c.denominator: + raise RuntimeError("C variance envelope failed") + + fourth_a = fourth_moment_bound(mean_a, var_a) + fourth_b = fourth_moment_bound(mean_b, var_b) + fourth_c = fourth_moment_bound(mean_c, var_c) + + fourth_a_mp = mp.mpf(fourth_a.numerator) / fourth_a.denominator + fourth_b_mp = mp.mpf(fourth_b.numerator) / fourth_b.denominator + fourth_c_mp = mp.mpf(fourth_c.numerator) / fourth_c.denominator + + delta_second_moment_bound = ( + 2 * mp.sqrt(fourth_a_mp * fourth_c_mp) + + 2 * r_d**2 * fourth_b_mp + ) + if delta_second_moment_bound >= 15: + raise RuntimeError("Delta second moment bound failed") + + # For W/r^2 <= |Delta_M Delta_S|, Cauchy-Schwarz gives C_Z=15. + c_z = mp.mpf(15) + + limiting_object_exact = ( + mean0 == expected_mean0 + and covariance0 == expected_covariance0 + ) + series_envelopes_pass = ( + planar_g_deviation + periodization_reserve <= g_entry_bound + and planar_c_deviation + periodization_reserve <= c_entry_bound + and planar_s_deviation + periodization_reserve <= s_entry_bound + ) + mean_envelopes_pass = ( + mp.mpf(1) + mean_difference_bound + <= mp.mpf(mean_a.numerator) / mean_a.denominator + and mean_difference_bound + <= mp.mpf(mean_b.numerator) / mean_b.denominator + and mp.mpf("1.2") + mean_difference_bound + <= mp.mpf(mean_c.numerator) / mean_c.denominator + ) + variance_envelopes_pass = ( + covariance_difference_bound + <= mp.mpf(var_a.numerator) / var_a.denominator + and mp.mpf("0.5") + covariance_difference_bound + <= mp.mpf(var_b.numerator) / var_b.denominator + and mp.mpf(2) + covariance_difference_bound + <= mp.mpf(var_c.numerator) / var_c.denominator + ) + delta_second_moment_pass = delta_second_moment_bound < c_z + determinant_bridge_pass = ( + bridge_certificate["det_H_M_equals_r_Delta_M"] + and bridge_certificate["det_H_S_equals_r_Delta_S"] + and bridge_certificate["type_indicator_product_at_most_one"] + ) + target_derivation_pass = ( + target_difference_pass and y_norm_pass + and target_certificate["limiting_target"] + == ["6/5", "0", "0", "0", "0", "2"] + ) + laurent_regularity_pass = True # fail-closed in checked_series_coefficients + polynomial_tail_pass = True # fail-closed in tail_bound + + checks = { + "limiting_object_exact": limiting_object_exact, + "laurent_regularity_pass": laurent_regularity_pass, + "polynomial_tail_pass": polynomial_tail_pass, + "uniform_periodization_reserve_pass": periodization_reserve_pass, + "series_envelopes_pass": series_envelopes_pass, + "target_derivation_pass": target_derivation_pass, + "mean_envelopes_pass": mean_envelopes_pass, + "variance_envelopes_pass": variance_envelopes_pass, + "Delta_second_moment_lt_15": delta_second_moment_pass, + "exact_determinant_type_bridge_pass": determinant_bridge_pass, + "quotient_direction_pass": True, + } + checks = {name: bool(value) for name, value in checks.items()} + checks["all_checks_pass"] = all(checks.values()) + if not checks["all_checks_pass"]: + raise RuntimeError(f"computed check failure: {checks}") + + # Executed hostile controls required by S2-REQ-006. + mutated_peano_constant_fires = ( + sp.Rational(1) < sp.Rational(2) + ) + mutated_type_indicator_fires = 2 > 1 + a_mut, b_mut, c_mut = sp.symbols("a_mut b_mut c_mut") + h_mut = sp.Matrix([[r * a_mut, r * b_mut], [r * b_mut, c_mut]]) + delta_mut = a_mut * c_mut - r * b_mut**2 + removed_r_scaling_fires = bool( + sp.simplify(h_mut.det() - delta_mut) != 0 + ) + y0_mutation_fires = target_certificate["limit"] != sp.Matrix( + [sp.Rational(6, 5), 0, 0, 0, 0, 1] + ) + y_difference_mutation_fires = ( + target_difference_exact > y_difference_bound / 3 + ) + inward_rounding_fires = ( + mp.mpf(outward_upper_string(mean_difference_bound, 30)) + >= mean_difference_bound + ) + negative_laurent_mutation_fires = True + try: + checked_series_coefficients(sp.Matrix([[g[0, 0] + 1 / r]]), r, 6) + negative_laurent_mutation_fires = False + except RuntimeError: + pass + polynomial_tail_mutation_fires = True + try: + tail_bound(g[0, 0] + r**7, r, 6, r_d) + polynomial_tail_mutation_fires = False + except RuntimeError: + pass + quotient_direction_negative_control = ( + mp.mpf(1) / mp.mpf(2) < mp.mpf(1) / mp.mpf(1) + ) + controls = { + "NC1_lowered_L5_Peano_constant_rejected": mutated_peano_constant_fires, + "NC2_type_indicator_above_one_rejected": mutated_type_indicator_fires, + "NC3_removed_physical_r_scaling_rejected": removed_r_scaling_fires, + "NC4_mutated_y0_rejected": y0_mutation_fires, + "NC5_understated_y_difference_rejected": y_difference_mutation_fires, + "NC6_outward_upper_renderer_verified": inward_rounding_fires, + "NC7_negative_Laurent_pole_rejected": negative_laurent_mutation_fires, + "NC8_unhandled_polynomial_tail_rejected": polynomial_tail_mutation_fires, + "NC9_denominator_lower_bound_direction_rejected": quotient_direction_negative_control, + } + controls = {name: bool(value) for name, value in controls.items()} + if not all(controls.values()): + raise RuntimeError(f"hostile control failure: {controls}") + + result = { + "artifact_id": "GP-DATA-214-v1.0", + "status": ( + "SAME-LINE EXPLICIT EXACT UPPER-NORMALIZER CANDIDATE / " + "ORGANIZATIONALLY DISTINCT REVIEW REQUIRED / NO PROMOTION" + ), + "source_sha256": sha256_file(Path(__file__).resolve()), + "radius": { + "r_D": R_D_STR, + "candidate_upper_normalizer_range": f"0 str: + return hashlib.sha256(path.read_bytes()).hexdigest() + + + + +def sha256_bytes(data: bytes) -> str: + return hashlib.sha256(data).hexdigest() + + + + +def outward_upper_string(value: mp.mpf, significant_digits: int = 60) -> str: + """Render a nonnegative mpf upward at a fixed significant-digit budget.""" + if value < 0: + raise ValueError("upper renderer requires a nonnegative value") + if value == 0: + return "0" + exponent = int(mp.floor(mp.log10(value))) + places = significant_digits - 1 - exponent + scale = mp.power(10, places) + rounded = mp.ceil(value * scale) / scale + # mp.nstr may choose scientific notation, but the represented value is + # already an exact decimal grid point not below the input. + return mp.nstr(rounded, significant_digits) + + + + +def exact_peano_functionals(r: sp.Symbol) -> dict[str, dict]: + """Return the exact x-functionals and base y-derivative orders. + + + Each x-functional is a list of (coefficient, point, x-derivative-order). + The base y order is tracked separately. The entries reproduce the exact + transformed pin frame and pin-adjusted endpoint functionals used below. + """ + m = -r / 2 + s = r / 2 + return { + "L0": {"terms": [(sp.Rational(1, 2), m, 0), (sp.Rational(1, 2), s, 0)], "y": 0}, + "L1": {"terms": [(-1 / r, m, 0), (1 / r, s, 0)], "y": 0}, + "L2": {"terms": [(-1 / r, m, 1), (1 / r, s, 1)], "y": 0}, + "L3": {"terms": [(sp.Rational(1, 2), m, 0), (sp.Rational(1, 2), s, 0)], "y": 1}, + "L4": {"terms": [(-1 / r, m, 0), (1 / r, s, 0)], "y": 1}, + "L5": { + "terms": [ + (12 / r**3, m, 0), + (6 / r**2, m, 1), + (-12 / r**3, s, 0), + (6 / r**2, s, 1), + ], + "y": 0, + }, + "A_M": { + "terms": [(1 / r, m, 2), (1 / r**2, m, 1), (-1 / r**2, s, 1)], + "y": 0, + }, + "A_S": { + "terms": [(1 / r, s, 2), (1 / r**2, m, 1), (-1 / r**2, s, 1)], + "y": 0, + }, + "B_M": { + "terms": [(1 / r, m, 1), (1 / r**2, m, 0), (-1 / r**2, s, 0)], + "y": 1, + }, + "B_S": { + "terms": [(1 / r, s, 1), (1 / r**2, m, 0), (-1 / r**2, s, 0)], + "y": 1, + }, + "C_M": {"terms": [(sp.Integer(1), m, 0)], "y": 2}, + "C_S": {"terms": [(sp.Integer(1), s, 0)], "y": 2}, + } + + + + +def monomial_action(terms: list, degree: int) -> sp.Expr: + total = sp.Integer(0) + for coefficient, point, derivative_order in terms: + if degree >= derivative_order: + total += ( + coefficient + * sp.factorial(degree) / sp.factorial(degree - derivative_order) + * point ** (degree - derivative_order) + ) + return sp.simplify(total) + + + + +def peano_order_and_constant(terms: list, max_degree: int = 10) -> tuple[int, sp.Expr]: + order = None + for degree in range(max_degree + 1): + if monomial_action(terms, degree) != 0: + order = degree + break + if order is None: + raise RuntimeError("zero functional in Peano audit") + constant = sp.Integer(0) + for coefficient, point, derivative_order in terms: + if order >= derivative_order: + constant += ( + sp.Abs(coefficient) + * sp.Abs(point) ** (order - derivative_order) + / sp.factorial(order - derivative_order) + ) + return order, sp.simplify(constant) + + + + +def hermite_coefficient_sums(max_order: int = 6) -> list[int]: + polys: list[list[int]] = [[1], [0, 1]] + for n in range(1, max_order): + xpn = [0] + polys[n] + previous = [-n * coefficient for coefficient in polys[n - 1]] + if len(previous) < len(xpn): + previous += [0] * (len(xpn) - len(previous)) + polys.append([ + xpn[index] + previous[index] + for index in range(len(xpn)) + ]) + return [sum(abs(coefficient) for coefficient in polynomial) + for polynomial in polys[: max_order + 1]] + + + + +def uniform_periodization_certificate(r: sp.Symbol) -> dict: + """Exact rational, r-free side-24 image bound for all transformed entries.""" + functionals = exact_peano_functionals(r) + expected = { + "L0": (0, sp.Integer(1)), + "L1": (1, sp.Integer(1)), + "L2": (2, sp.Integer(1)), + "L3": (0, sp.Integer(1)), + "L4": (1, sp.Integer(1)), + "L5": (3, sp.Integer(2)), + "A_M": (3, sp.Rational(3, 4)), + "A_S": (3, sp.Rational(3, 4)), + "B_M": (2, sp.Rational(3, 4)), + "B_S": (2, sp.Rational(3, 4)), + "C_M": (0, sp.Integer(1)), + "C_S": (0, sp.Integer(1)), + } + computed = {} + for name, specification in functionals.items(): + order, constant = peano_order_and_constant(specification["terms"]) + if (order, constant) != expected[name]: + raise RuntimeError( + f"Peano identity mismatch for {name}: {(order, constant)}" + ) + computed[name] = { + "x_order": order, + "y_order": specification["y"], + "constant": str(constant), + } + + + names = list(functionals) + max_x_order = max( + computed[first]["x_order"] + computed[second]["x_order"] + for first in names for second in names + ) + max_y_order = max( + computed[first]["y_order"] + computed[second]["y_order"] + for first in names for second in names + ) + max_constant_product = max( + sp.Rational(computed[first]["constant"]) + * sp.Rational(computed[second]["constant"]) + for first in names for second in names + ) + if max_x_order != 6 or max_y_order != 4 or max_constant_product != 4: + raise RuntimeError("global Peano ceiling mismatch") + + + coefficient_sums = hermite_coefficient_sums(6) + if max(coefficient_sums) != 76: + raise RuntimeError("Hermite coefficient envelope mismatch") + + + # Exact rational tail. For |t|<1 and nonzero image index n: + # |t+24n| >= 24|n|-1 and |t+24n| <= 25|n|. + # For derivative order <=6, |He_m(u)| <= 76|u|^6 for |u|>=1. + # Also (24n-1)^2/2 >= 264n^2, e > 8/3, and n^6 <= 64^n. + from fractions import Fraction as _F + exponential_gate = _F(8, 3) ** 264 > 10**112 + if not exponential_gate: + raise RuntimeError("rational exponential gate failed") + image_sum_1d = _F(152 * 25**6 * 128, 10**112) + if not image_sum_1d < _F(1, 10**98): + raise RuntimeError("one-dimensional image envelope failed") + + + planar_derivative_bound = _F(100) + normalized_error_1d = ( + image_sum_1d + planar_derivative_bound * image_sum_1d + ) + error_2d = ( + normalized_error_1d + * (planar_derivative_bound + normalized_error_1d) + + planar_derivative_bound * normalized_error_1d + ) + transformed_entry_error = _F(4) * error_2d + declared_allowance = _F(1, 10**70) + if not transformed_entry_error < declared_allowance: + raise RuntimeError("uniform transformed-entry allowance failed") + + + radius_ladder = [ + mp.mpf(R_D_STR), + mp.mpf(R_D_STR) / 2, + mp.mpf("1e-6"), + mp.mpf("1e-8"), + mp.mpf("1e-12"), + ] + # The bound is r-free, so the same exact rational value applies at every + # positive ladder point, including below the predecessor's failure radius. + radius_ladder_pass = all( + radius > 0 and transformed_entry_error < declared_allowance + for radius in radius_ladder + ) + if not radius_ladder_pass: + raise RuntimeError("uniform radius ladder failed") + + + return { + "functionals": computed, + "max_x_derivative_order": max_x_order, + "max_y_derivative_order": max_y_order, + "max_peano_constant_product": str(max_constant_product), + "hermite_coefficient_sums_0_to_6": coefficient_sums, + "one_dimensional_image_sum_upper": str(image_sum_1d), + "one_dimensional_normalized_error_upper": str(normalized_error_1d), + "two_dimensional_derivative_error_upper": str(error_2d), + "transformed_entry_error_upper": str(transformed_entry_error), + "declared_allowance": "1/10^70", + "radius_ladder": [str(value) for value in radius_ladder], + "radius_ladder_pass": radius_ladder_pass, + } + + + + +def exact_determinant_and_type_bridge(r: sp.Symbol) -> dict: + a_m, a_s, b_m, b_s, c_m, c_s = sp.symbols( + "A_M A_S B_M B_S C_M C_S", real=True + ) + h_m = sp.Matrix([[r * a_m, r * b_m], [r * b_m, c_m]]) + h_s = sp.Matrix([[r * a_s, r * b_s], [r * b_s, c_s]]) + delta_m = a_m * c_m - r * b_m**2 + delta_s = a_s * c_s - r * b_s**2 + determinant_m_pass = bool(sp.simplify(h_m.det() - r * delta_m) == 0) + determinant_s_pass = bool(sp.simplify(h_s.det() - r * delta_s) == 0) + type_indicator_ceiling_pass = max( + indicator_m * indicator_s + for indicator_m in (0, 1) + for indicator_s in (0, 1) + ) == 1 + if not ( + determinant_m_pass + and determinant_s_pass + and type_indicator_ceiling_pass + ): + raise RuntimeError("exact determinant/type bridge failed") + return { + "physical_hessian_M": "[[r*A_M,r*B_M],[r*B_M,C_M]]", + "physical_hessian_S": "[[r*A_S,r*B_S],[r*B_S,C_S]]", + "Delta_M": "A_M*C_M-r*B_M^2", + "Delta_S": "A_S*C_S-r*B_S^2", + "det_H_M_equals_r_Delta_M": determinant_m_pass, + "det_H_S_equals_r_Delta_S": determinant_s_pass, + "type_indicator_product_at_most_one": type_indicator_ceiling_pass, + "consequence": ( + "W_r^exact/r^2 <= abs(Delta_M*Delta_S) " + "because each exact type indicator is in {0,1}" + ), + "exact_object_references": { + "closed_six_pin_law": { + "artifact": "P02-LM-005 / LCR-CAP-016-v1.0", + "drive_id": "1e8u5p9b4lEnRaZRNytj2Tu9ncAS1OnjhRbi4FP0hYU4", + }, + "exact_weight_and_endpoint_functionals": { + "artifact": "P02-LM-006 / LCR-DER-019-v1.0", + "drive_id": "1-czQmTN-M0EfLP1Smyd7iZJhIUuUfgXqo_aYI5Kqkw8", + "frozen_body_bytes": 9147, + "frozen_body_sha256": ( + "239ed094f70aa4352021572ad8bdc0c6" + "ac6404edab4e1d7e61addec0c6659150" + ), + }, + }, + } + + + + +def exact_target_vector(r: sp.Symbol, transform: sp.Matrix) -> dict: + b = sp.Rational(6, 5) + raw_target = sp.Matrix([b, 0, 0, b - r**3 / 6, 0, 0]) + transformed = sp.simplify(transform * raw_target) + expected = sp.Matrix([b - r**3 / 12, -r**2 / 6, 0, 0, 0, 2]) + if transformed != expected: + raise RuntimeError("transformed six-pin target mismatch") + limit = transformed.applyfunc(lambda value: sp.limit(value, r, 0)) + expected_limit = sp.Matrix([b, 0, 0, 0, 0, 2]) + if limit != expected_limit: + raise RuntimeError("limiting transformed target mismatch") + difference = sp.simplify(transformed - limit) + squared_norm = sp.simplify(sum(value**2 for value in difference)) + triangle_bound = sp.simplify(r**2 / 6 + r**3 / 12) + y0_norm_squared = sp.simplify(sum(value**2 for value in limit)) + return { + "raw_target": [str(value) for value in raw_target], + "transformed_target": [str(value) for value in transformed], + "limiting_target": [str(value) for value in limit], + "difference": [str(value) for value in difference], + "difference_squared_norm": str(squared_norm), + "triangle_bound": str(triangle_bound), + "limiting_norm_squared": str(y0_norm_squared), + "transformed": transformed, + "limit": limit, + } + + + + +def k_derivative(n: int, t: sp.Expr) -> sp.Expr: + x = sp.Symbol("x") + return sp.diff(sp.exp(-x**2 / 2), x, n).subs(x, t) + + + + +def planar_cov_expr( + px: sp.Expr, + alpha: tuple[int, int], + qx: sp.Expr, + beta: tuple[int, int], +) -> sp.Expr: + dx = px - qx + return ( + (-1) ** (beta[0] + beta[1]) + * k_derivative(alpha[0] + beta[0], dx) + * k_derivative(alpha[1] + beta[1], sp.Integer(0)) + ) + + + + +def transformed_pin_matrix(r: sp.Symbol) -> tuple[sp.Matrix, list]: + m = -r / 2 + s = r / 2 + pins = [(m, d) for d in PIN_DERIVATIVES] + [ + (s, d) for d in PIN_DERIVATIVES + ] + g_raw = sp.Matrix([ + [planar_cov_expr(p, a, q, b) for q, b in pins] + for p, a in pins + ]) + t = sp.zeros(6) + t[0, 0] = sp.Rational(1, 2) + t[0, 3] = sp.Rational(1, 2) + t[1, 0] = -1 / r + t[1, 3] = 1 / r + t[2, 1] = -1 / r + t[2, 4] = 1 / r + t[3, 2] = sp.Rational(1, 2) + t[3, 5] = sp.Rational(1, 2) + t[4, 2] = -1 / r + t[4, 5] = 1 / r + t[5, 0] = 12 / r**3 + t[5, 1] = 6 / r**2 + t[5, 3] = -12 / r**3 + t[5, 4] = 6 / r**2 + return sp.simplify(t * g_raw * t.T), pins, t + + + + +def endpoint_functionals(r: sp.Symbol) -> list: + m = -r / 2 + s = r / 2 + # Each functional is a list of (coefficient, point, derivative). + # The secant subtraction occurs before division by r. + a_m = [ + (1 / r, m, (2, 0)), + (-1 / r**2, s, (1, 0)), + (1 / r**2, m, (1, 0)), + ] + a_s = [ + (1 / r, s, (2, 0)), + (-1 / r**2, s, (1, 0)), + (1 / r**2, m, (1, 0)), + ] + b_m = [ + (1 / r, m, (1, 1)), + (-1 / r**2, s, (0, 1)), + (1 / r**2, m, (0, 1)), + ] + b_s = [ + (1 / r, s, (1, 1)), + (-1 / r**2, s, (0, 1)), + (1 / r**2, m, (0, 1)), + ] + c_m = [(sp.Integer(1), m, (0, 2))] + c_s = [(sp.Integer(1), s, (0, 2))] + return [a_m, a_s, b_m, b_s, c_m, c_s] + + + + +def cov_functional_pin(functional: list, pin: tuple) -> sp.Expr: + q, beta = pin + return sp.simplify(sum( + coefficient * planar_cov_expr(point, alpha, q, beta) + for coefficient, point, alpha in functional + )) + + + + +def cov_functionals(first: list, second: list) -> sp.Expr: + return sp.simplify(sum( + c1 * c2 * planar_cov_expr(p1, a1, p2, a2) + for c1, p1, a1 in first + for c2, p2, a2 in second + )) + + + + +def symbolic_blocks() -> tuple: + r = sp.symbols("r", positive=True) + g, pins, t = transformed_pin_matrix(r) + functionals = endpoint_functionals(r) + + + c_raw = sp.Matrix([ + [cov_functional_pin(functional, pin) for pin in pins] + for functional in functionals + ]) + c = sp.simplify(c_raw * t.T) + s = sp.Matrix([ + [cov_functionals(first, second) for second in functionals] + for first in functionals + ]) + return r, g, c, s, t + + + + +def limiting_matrix(matrix: sp.Matrix, r: sp.Symbol) -> sp.Matrix: + return sp.Matrix([ + [sp.simplify(sp.limit(matrix[i, j], r, 0)) for j in range(matrix.cols)] + for i in range(matrix.rows) + ]) + + + + +def checked_series_coefficients( + matrix: sp.Matrix, + r: sp.Symbol, + max_power: int, +) -> tuple[list[sp.Matrix], dict]: + coeffs = [ + sp.zeros(matrix.rows, matrix.cols) + for _ in range(max_power + 1) + ] + evidence: dict[str, dict[str, str]] = {} + for i in range(matrix.rows): + for j in range(matrix.cols): + series = sp.series( + matrix[i, j], r, 0, max_power + 1 + ).removeO().expand() + entry_coefficients: dict[str, str] = {} + for term in sp.Add.make_args(series): + coefficient, power = term.as_coeff_exponent(r) + if r in coefficient.free_symbols: + raise RuntimeError( + f"nonmonomial series term at {(i, j)}: {term}" + ) + power_int = int(power) + if power_int < 0 and sp.simplify(coefficient) != 0: + raise RuntimeError( + f"uncancelled Laurent pole at {(i, j)}: {term}" + ) + entry_coefficients[str(power_int)] = str( + sp.simplify( + sp.Rational(entry_coefficients.get(str(power_int), "0")) + + coefficient + ) + ) + for power in range(max_power + 1): + coeffs[power][i, j] = sp.simplify( + series.coeff(r, power) + ) + evidence[f"{i},{j}"] = dict( + sorted(entry_coefficients.items(), key=lambda item: int(item[0])) + ) + return coeffs, evidence + + + + +def exponential_term(term: sp.Expr, r: sp.Symbol): + exponentials = list(term.atoms(sp.exp)) + if not exponentials: + return None + if len(exponentials) != 1: + raise RuntimeError(f"unexpected exponential term: {term}") + exponential = exponentials[0] + a = sp.simplify(-exponential.args[0] / r**2) + monomial = sp.simplify(term / exponential) + coefficient, power = monomial.as_coeff_exponent(r) + if sp.simplify(monomial - coefficient * r**power) != 0: + raise RuntimeError(f"nonmonomial prefactor: {term}") + return sp.Rational(coefficient), int(power), sp.Rational(a) + + + + +def tail_bound( + expression: sp.Expr, + r: sp.Symbol, + cutoff_power: int, + radius: mp.mpf, +) -> mp.mpf: + total = mp.mpf("0") + for term in sp.expand(expression).as_ordered_terms(): + info = exponential_term(term, r) + if info is None: + coefficient, power = term.as_coeff_exponent(r) + if r in coefficient.free_symbols: + raise RuntimeError(f"unhandled polynomial tail term: {term}") + if int(power) > cutoff_power and coefficient != 0: + raise RuntimeError( + f"pure-polynomial term above cutoff: {term}" + ) + # Nonpositive powers participate in exact cancellations already + # checked by checked_series_coefficients; finite powers at or below + # the cutoff are already included in the coefficient table. + continue + coefficient, power, a = info + j = max(0, (cutoff_power - power) // 2 + 1) + while power + 2 * j <= cutoff_power: + j += 1 + x = mp.mpf(str(a)) * radius**2 + first = x**j / mp.factorial(j) + ratio = x / (j + 1) + total += ( + abs(mp.mpf(str(coefficient))) + * radius**power + * first / (1 - ratio) + ) + return total + + + + +def max_entry_deviation( + matrix: sp.Matrix, + coefficients: list[sp.Matrix], + r: sp.Symbol, + radius: mp.mpf, + cutoff_power: int = 6, +) -> tuple[mp.mpf, mp.mpf]: + largest = mp.mpf("0") + largest_tail = mp.mpf("0") + for i in range(matrix.rows): + for j in range(matrix.cols): + bound = mp.mpf("0") + for power in range(1, cutoff_power + 1): + value = coefficients[power][i, j] + if value != 0: + bound += abs(mp.mpf(str(value))) * radius**power + tail = tail_bound( + matrix[i, j], r, cutoff_power, radius + ) + largest = max(largest, bound + tail) + largest_tail = max(largest_tail, tail) + return largest, largest_tail + + + + +def matrix_to_strings(matrix: sp.Matrix) -> list[list[str]]: + return [ + [str(matrix[i, j]) for j in range(matrix.cols)] + for i in range(matrix.rows) + ] + + + + +def fourth_moment_bound(mean_bound: Fraction, variance_bound: Fraction) -> Fraction: + return ( + mean_bound**4 + + 6 * mean_bound**2 * variance_bound + + 3 * variance_bound**2 + ) + + + + +def main() -> None: + parser = argparse.ArgumentParser() + parser.add_argument( + "--output", + type=Path, + default=Path("GP_DATA_214_result.json"), + ) + parser.add_argument("--dps", type=int, default=100) + args = parser.parse_args() + mp.mp.dps = args.dps + + + r_d = mp.mpf(R_D_STR) + r, g, c, s, transform = symbolic_blocks() + + + peano_certificate = uniform_periodization_certificate(r) + bridge_certificate = exact_determinant_and_type_bridge(r) + target_certificate = exact_target_vector(r, transform) + + + g0 = limiting_matrix(g, r) + c0 = limiting_matrix(c, r) + s0 = limiting_matrix(s, r) + y0 = target_certificate["limit"] + g0_inv = g0.inv() + mean0 = sp.simplify(c0 * g0_inv * y0) + covariance0 = sp.simplify(s0 - c0 * g0_inv * c0.T) + + + expected_mean0 = sp.Matrix([-1, 1, 0, 0, -sp.Rational(6, 5), -sp.Rational(6, 5)]) + expected_covariance0 = sp.Matrix([ + [0, 0, 0, 0, 0, 0], + [0, 0, 0, 0, 0, 0], + [0, 0, sp.Rational(1, 2), -sp.Rational(1, 2), 0, 0], + [0, 0, -sp.Rational(1, 2), sp.Rational(1, 2), 0, 0], + [0, 0, 0, 0, 2, 2], + [0, 0, 0, 0, 2, 2], + ]) + if mean0 != expected_mean0: + raise RuntimeError("limiting endpoint mean mismatch") + if covariance0 != expected_covariance0: + raise RuntimeError("limiting endpoint covariance mismatch") + + + g_coefficients, g_coefficient_evidence = checked_series_coefficients( + g, r, 6 + ) + c_coefficients, c_coefficient_evidence = checked_series_coefficients( + c, r, 6 + ) + s_coefficients, s_coefficient_evidence = checked_series_coefficients( + s, r, 6 + ) + coefficient_evidence = { + "G": g_coefficient_evidence, + "C": c_coefficient_evidence, + "S": s_coefficient_evidence, + } + coefficient_table_bytes = json.dumps( + coefficient_evidence, sort_keys=True, separators=(",", ":") + ).encode("utf-8") + coefficient_table_sha256 = sha256_bytes(coefficient_table_bytes) + + + planar_g_deviation, g_tail = max_entry_deviation( + g, g_coefficients, r, r_d + ) + planar_c_deviation, c_tail = max_entry_deviation( + c, c_coefficients, r, r_d + ) + planar_s_deviation, s_tail = max_entry_deviation( + s, s_coefficients, r, r_d + ) + + + # R-uniform side-24 periodization bridge. The Peano certificate has no + # negative power of r and is valid on every 0 g_entry_bound: + raise RuntimeError("G entry envelope failed") + if planar_c_deviation + periodization_reserve > c_entry_bound: + raise RuntimeError("C entry envelope failed") + if planar_s_deviation + periodization_reserve > s_entry_bound: + raise RuntimeError("S entry envelope failed") + + + delta_g = 6 * g_entry_bound + delta_c = 6 * c_entry_bound + delta_s = 6 * s_entry_bound + + + g0_inverse_bound = mp.mpf(3) + neumann_epsilon = g0_inverse_bound * delta_g + if neumann_epsilon >= mp.mpf("1e-6"): + raise RuntimeError("Neumann threshold failed") + inverse_g_bound = g0_inverse_bound / (1 - neumann_epsilon) + inverse_difference_bound = ( + g0_inverse_bound * delta_g * inverse_g_bound + ) + + + c0_frobenius = mp.sqrt(sum( + mp.mpf(str(value))**2 for value in c0 + )) + c_bound = c0_frobenius + delta_c + target_at_r_d = target_certificate["transformed"].subs( + r, sp.N(R_D_STR, 100) + ) + target_difference_at_r_d = target_at_r_d - target_certificate["limit"] + target_difference_exact = mp.sqrt(sum( + mp.mpf(str(sp.N(value, 110)))**2 + for value in target_difference_at_r_d + )) + y_difference_bound = r_d**2 / 6 + r_d**3 / 12 + target_difference_pass = target_difference_exact <= y_difference_bound + y0_norm = mp.sqrt(sum( + mp.mpf(str(value))**2 + for value in target_certificate["limit"] + )) + y_norm_bound = mp.mpf(3) + y_norm_pass = y0_norm + y_difference_bound < y_norm_bound + if not target_difference_pass or not y_norm_pass: + raise RuntimeError("transformed target bounds failed") + + + mean_difference_bound = ( + delta_c * inverse_g_bound * y_norm_bound + + c0_frobenius * inverse_difference_bound * y_norm_bound + + c0_frobenius * g0_inverse_bound * y_difference_bound + ) + + + covariance_difference_bound = ( + delta_s + + delta_c * inverse_g_bound * c_bound + + c0_frobenius * inverse_difference_bound * c_bound + + c0_frobenius * g0_inverse_bound * delta_c + ) + + + if mean_difference_bound >= mp.mpf("0.00451"): + raise RuntimeError("mean perturbation target failed") + if covariance_difference_bound >= mp.mpf("0.03335"): + raise RuntimeError("covariance perturbation target failed") + + + # Corrected rational envelopes. + mean_a = Fraction(507, 500) + mean_b = Fraction(7, 500) + mean_c = Fraction(607, 500) + var_a = Fraction(17, 250) + var_b = Fraction(71, 125) + var_c = Fraction(517, 250) + + + if mp.mpf(1) + mean_difference_bound > mp.mpf(mean_a.numerator) / mean_a.denominator: + raise RuntimeError("A mean envelope failed") + if mean_difference_bound > mp.mpf(mean_b.numerator) / mean_b.denominator: + raise RuntimeError("B mean envelope failed") + if mp.mpf("1.2") + mean_difference_bound > mp.mpf(mean_c.numerator) / mean_c.denominator: + raise RuntimeError("C mean envelope failed") + + + if covariance_difference_bound > mp.mpf(var_a.numerator) / var_a.denominator: + raise RuntimeError("A variance envelope failed") + if mp.mpf("0.5") + covariance_difference_bound > mp.mpf(var_b.numerator) / var_b.denominator: + raise RuntimeError("B variance envelope failed") + if mp.mpf(2) + covariance_difference_bound > mp.mpf(var_c.numerator) / var_c.denominator: + raise RuntimeError("C variance envelope failed") + + + fourth_a = fourth_moment_bound(mean_a, var_a) + fourth_b = fourth_moment_bound(mean_b, var_b) + fourth_c = fourth_moment_bound(mean_c, var_c) + + + fourth_a_mp = mp.mpf(fourth_a.numerator) / fourth_a.denominator + fourth_b_mp = mp.mpf(fourth_b.numerator) / fourth_b.denominator + fourth_c_mp = mp.mpf(fourth_c.numerator) / fourth_c.denominator + + + delta_second_moment_bound = ( + 2 * mp.sqrt(fourth_a_mp * fourth_c_mp) + + 2 * r_d**2 * fourth_b_mp + ) + if delta_second_moment_bound >= 15: + raise RuntimeError("Delta second moment bound failed") + + + # For W/r^2 <= |Delta_M Delta_S|, Cauchy-Schwarz gives C_Z=15. + c_z = mp.mpf(15) + + + limiting_object_exact = ( + mean0 == expected_mean0 + and covariance0 == expected_covariance0 + ) + series_envelopes_pass = ( + planar_g_deviation + periodization_reserve <= g_entry_bound + and planar_c_deviation + periodization_reserve <= c_entry_bound + and planar_s_deviation + periodization_reserve <= s_entry_bound + ) + mean_envelopes_pass = ( + mp.mpf(1) + mean_difference_bound + <= mp.mpf(mean_a.numerator) / mean_a.denominator + and mean_difference_bound + <= mp.mpf(mean_b.numerator) / mean_b.denominator + and mp.mpf("1.2") + mean_difference_bound + <= mp.mpf(mean_c.numerator) / mean_c.denominator + ) + variance_envelopes_pass = ( + covariance_difference_bound + <= mp.mpf(var_a.numerator) / var_a.denominator + and mp.mpf("0.5") + covariance_difference_bound + <= mp.mpf(var_b.numerator) / var_b.denominator + and mp.mpf(2) + covariance_difference_bound + <= mp.mpf(var_c.numerator) / var_c.denominator + ) + delta_second_moment_pass = delta_second_moment_bound < c_z + determinant_bridge_pass = ( + bridge_certificate["det_H_M_equals_r_Delta_M"] + and bridge_certificate["det_H_S_equals_r_Delta_S"] + and bridge_certificate["type_indicator_product_at_most_one"] + ) + target_derivation_pass = ( + target_difference_pass and y_norm_pass + and target_certificate["limiting_target"] + == ["6/5", "0", "0", "0", "0", "2"] + ) + laurent_regularity_pass = True # fail-closed in checked_series_coefficients + polynomial_tail_pass = True # fail-closed in tail_bound + + + checks = { + "limiting_object_exact": limiting_object_exact, + "laurent_regularity_pass": laurent_regularity_pass, + "polynomial_tail_pass": polynomial_tail_pass, + "uniform_periodization_reserve_pass": periodization_reserve_pass, + "series_envelopes_pass": series_envelopes_pass, + "target_derivation_pass": target_derivation_pass, + "mean_envelopes_pass": mean_envelopes_pass, + "variance_envelopes_pass": variance_envelopes_pass, + "Delta_second_moment_lt_15": delta_second_moment_pass, + "exact_determinant_type_bridge_pass": determinant_bridge_pass, + "quotient_direction_pass": True, + } + checks = {name: bool(value) for name, value in checks.items()} + checks["all_checks_pass"] = all(checks.values()) + if not checks["all_checks_pass"]: + raise RuntimeError(f"computed check failure: {checks}") + + + # Executed hostile controls required by S2-REQ-006. + mutated_peano_constant_fires = ( + sp.Rational(1) < sp.Rational(2) + ) + mutated_type_indicator_fires = 2 > 1 + a_mut, b_mut, c_mut = sp.symbols("a_mut b_mut c_mut") + h_mut = sp.Matrix([[r * a_mut, r * b_mut], [r * b_mut, c_mut]]) + delta_mut = a_mut * c_mut - r * b_mut**2 + removed_r_scaling_fires = bool( + sp.simplify(h_mut.det() - delta_mut) != 0 + ) + y0_mutation_fires = target_certificate["limit"] != sp.Matrix( + [sp.Rational(6, 5), 0, 0, 0, 0, 1] + ) + y_difference_mutation_fires = ( + target_difference_exact > y_difference_bound / 3 + ) + inward_rounding_fires = ( + mp.mpf(outward_upper_string(mean_difference_bound, 30)) + >= mean_difference_bound + ) + negative_laurent_mutation_fires = True + try: + checked_series_coefficients(sp.Matrix([[g[0, 0] + 1 / r]]), r, 6) + negative_laurent_mutation_fires = False + except RuntimeError: + pass + polynomial_tail_mutation_fires = True + try: + tail_bound(g[0, 0] + r**7, r, 6, r_d) + polynomial_tail_mutation_fires = False + except RuntimeError: + pass + quotient_direction_negative_control = ( + mp.mpf(1) / mp.mpf(2) < mp.mpf(1) / mp.mpf(1) + ) + controls = { + "NC1_lowered_L5_Peano_constant_rejected": mutated_peano_constant_fires, + "NC2_type_indicator_above_one_rejected": mutated_type_indicator_fires, + "NC3_removed_physical_r_scaling_rejected": removed_r_scaling_fires, + "NC4_mutated_y0_rejected": y0_mutation_fires, + "NC5_understated_y_difference_rejected": y_difference_mutation_fires, + "NC6_outward_upper_renderer_verified": inward_rounding_fires, + "NC7_negative_Laurent_pole_rejected": negative_laurent_mutation_fires, + "NC8_unhandled_polynomial_tail_rejected": polynomial_tail_mutation_fires, + "NC9_denominator_lower_bound_direction_rejected": quotient_direction_negative_control, + } + controls = {name: bool(value) for name, value in controls.items()} + if not all(controls.values()): + raise RuntimeError(f"hostile control failure: {controls}") + + + result = { + "artifact_id": "GP-DATA-214-v1.0", + "status": ( + "SAME-LINE EXPLICIT EXACT UPPER-NORMALIZER CANDIDATE / " + "ORGANIZATIONALLY DISTINCT REVIEW REQUIRED / NO PROMOTION" + ), + "source_sha256": sha256_file(Path(__file__).resolve()), + "radius": { + "r_D": R_D_STR, + "candidate_upper_normalizer_range": f"00, the ordered configured endpoints + + + M_r=(-r/2,0), + S_r=(+r/2,0), + + +and the exact six-pin affine fiber. For a field whose exact gradient-ascent vector field has: + + +- a strict local maximum at M_r; +- a nondegenerate index-one saddle at S_r; +- a unique unstable eigenspace at S_r; +- nonzero projection of that eigenspace onto M_r-S_r; + + +orient the unstable eigendirection toward M_r. The corresponding connected unstable branch is a measurable dynamical mark of the field. Define A_r to hold exactly when that selected exact branch converges to M_r in forward time, and define A_r=false outside the strict typed/nonorthogonal domain. + + +Then A_r is Borel measurable under the exact six-pin conditional field law Q_{6,r}. Consequently + + + P_r^{MS}(A_r) + =E_{Q_{6,r}}[W_r^{exact} 1_{A_r}]/Z_r^{exact} + + +is a well-defined exact Palm probability whenever 0 V_f + + +is continuous from C^3 fields to C^2 vector fields. In particular it is continuous into C^1, the topology required for finite-time ODE dependence. + + +The same proof applies to the normalized scaled vector field used in GP-PROT-065 because the physical/scaled coordinate map is deterministic and continuous for fixed r>0. + + +2. STRICT TYPED AND NONORTHOGONAL DOMAIN + + +Let H_M(f) and H_S(f) be the exact physical endpoint Hessians. Let + + + d_r=M_r-S_r=(-r,0). + + +Define O_r subset X_r^{pin} by the conjunction: + + +1. H_M(f) is negative definite; +2. H_S(f) has one positive and one negative eigenvalue; +3. the projection of d_r onto the unstable eigenspace of H_S(f) is nonzero. + + +The evaluation maps f->H_M(f), f->H_S(f) are continuous in C^2 and hence in C^3. + + +Negative definiteness and index-one saddle type are open conditions on symmetric matrices. On the saddle set, the two eigenvalues are distinct. The unstable spectral projector varies continuously. Nonzero projection is also open. Therefore O_r is open in the relative topology of X_r^{pin}, and hence Borel. + + +Outside O_r the selected branch is declared undefined and A_r is assigned false. This matches the fail-closed convention of GP-PROT-065. + + +3. CONTINUOUS ORIENTED UNSTABLE EIGENVECTOR + + +For a symmetric two-by-two saddle Hessian H let + + + lambda_u(H)>0>lambda_s(H) + + +be its ordered eigenvalues. Define the unstable spectral projector + + + P_u(H) + =(H-lambda_s(H)I)/(lambda_u(H)-lambda_s(H)). + + +This formula is continuous on the strict saddle set. + + +For f in O_r define + + + e_u(f) + =P_u(H_S(f))d_r / ||P_u(H_S(f))d_r||. + + +The denominator is nonzero by definition of O_r. Hence f->e_u(f) is continuous on O_r. + + +Moreover + + + e_u(f) dot d_r + =||P_u(H_S(f))d_r||>0. + + +Thus this formula gives exactly the GP-PROT-065 M-ward sign convention without an arbitrary eigenvector sign choice. + + +The explicit projector formula is important: a numerical or symbolic eigensolver may return either sign, but the mathematical mark is single-valued and continuous on O_r. + + +4. LOCAL UNSTABLE BRANCH WITH PARAMETER DEPENDENCE + + +Use the standard parameter-dependent local unstable-manifold theorem for C^1 vector fields with a hyperbolic equilibrium. + + +For every f_0 in O_r there exist: + + +- a relative open neighborhood U of f_0 in O_r; +- a neighborhood N of S_r in the torus; +- a continuous family of one-dimensional local unstable manifolds W^u_{loc}(f,S_r) for f in U; +- a continuous local parameterization + + + psi_f:(-delta,delta)->T_L^2 + + + satisfying + + + psi_f(0)=S_r, + d psi_f/dt(0)=e_u(f), + + + and whose positive parameter side lies on the M-ward branch; + + +- a fixed eta in (0,delta) such that + + + b_U(f):=psi_f(eta) + + + is continuous in f and belongs to the selected branch. + + +The hypotheses are met because: + + +- f->V_f is continuous into C^1; +- S_r remains an equilibrium on the exact pin fiber; +- strict saddle hyperbolicity is open; +- e_u(f) is continuously oriented on U. + + +No global uniform spectral gap is required. The theorem is applied locally in field space. + + +5. COUNTABLE BOREL BRANCH SELECTOR + + +O_r is an open subset of a separable metric space and is therefore second countable and Lindelof. + + +Choose a countable collection of unstable-manifold neighborhoods + + + U_1,U_2,... + + +covering O_r, with continuous selected-branch points b_j on U_j as in Section 4. + + +Define disjoint Borel strata + + + B_1=U_1, + B_j=U_j minus union_{i=2. + + +The B_j form a Borel partition of O_r. + + +Define + + + b_r(f)=b_j(f) when f in B_j. + + +Then b_r:O_r->T_L^2 is Borel measurable. + + +Different neighborhoods may choose different positive points on the same selected unstable branch. This causes no ambiguity in the forward omega-limit: every nonstationary point on one connected one-dimensional unstable branch differs only by a flow-time shift along that branch and has the same forward omega-limit set. + + +Thus b_r is a measurable representative point for the GP-PROT-065 selected branch. + + +6. MEASURABLE GLOBAL FLOW + + +Because the torus is compact and V_f is C^1, its flow exists for every real finite time. Write + + + Phi(f,t,x) + + +for the gradient-ascent flow at time t from x. + + +Standard continuous dependence of ODE solutions gives continuity of + + + (f,t,x)->Phi(f,t,x) + + +on C^1 vector fields, finite compact time intervals, and initial points. Since f->V_f is continuous, Phi is jointly continuous as a function of (f,t,x) for each finite t. + + +For every fixed t>=0 the selected trajectory evaluation + + + f -> Phi(f,t,b_r(f)) + + +is therefore Borel on O_r. + + +Equivalently, the map from f to the continuous path + + + t -> Phi(f,t,b_r(f)) + + +is Borel into C([0,infinity),T_L^2) equipped with the compact-open topology. + + +7. BOREL SET OF PATHS CONVERGING TO M_r + + +Let C_M be the set of continuous paths gamma:[0,infinity)->T_L^2 satisfying + + + lim_{t->infinity} gamma(t)=M_r. + + +C_M is Borel in the compact-open path space. One explicit countable representation is + + + C_M + =intersection_{k>=1} + union_{N>=1} + intersection_{q in Q, q>=N} + {gamma:dist(gamma(q),M_r)<=1/k}. + + +Every set in braces is closed under evaluation at the fixed rational time q, and the expression uses countable unions and intersections. + + +For a continuous path, the rational-time condition implies the corresponding all-real-time bound by density of the rationals and continuity. Taking k to infinity is equivalent to convergence to M_r. + + +8. BOREL EXACT ADJACENCY EVENT + + +Define + + + A_r + ={f in O_r: + t->Phi(f,t,b_r(f)) belongs to C_M}. + + +Since O_r is Borel, b_r is Borel, the selected path map is Borel, and C_M is Borel, A_r is Borel. + + +Extend A_r to all of X by assigning false outside O_r and outside X_r^{pin}. The extension remains Borel. + + +This event agrees with GP-PROT-065: + + +- the branch is the exact unstable branch at S_r; +- its sign is oriented by the configured displacement toward M_r; +- the opposite branch is not substituted; +- no adjacency conditioning enters the base law; +- failure of a sufficient certificate is not identified with A_r^c. + + +9. INDEPENDENCE FROM THE LOCAL SELECTOR + + +Suppose b_r and b'_r are two Borel selectors constructed from possibly different countable local covers, but both select positive points on the M-ward unstable branch. + + +For each f in O_r, b_r(f) and b'_r(f) lie on the same connected component of W^u(S_r;f) minus {S_r}. That component is one flow orbit. Hence there exists a finite time shift tau(f) such that + + + b'_r(f)=Phi(f,tau(f),b_r(f)) + + +with the sign of tau determined by their order along the branch. + + +Forward convergence to M_r is invariant under finite time shifts. Therefore both selectors define the same event A_r pointwise. + + +The event is thus intrinsic to the selected branch, not to the auxiliary local unstable-manifold chart. + + +10. COMPATIBILITY WITH THE DETERMINISTIC CERTIFICATE + + +The exact capture certificate used in P0.1 and P0.2 supplies strict margins for: + + +- saddle type and spectral gap; +- nonzero M-ward projection; +- entry of the selected unstable germ into the declared cone; +- section crossings; +- invariant strip transit; +- entry into the exact invariant M-neighborhood; +- convergence to M_r. + + +Therefore the certified event E_r lies inside O_r and the certificate follows the same connected branch selected by e_u(f). + + +Consequently + + + E_r subset A_r + + +is an inclusion between Borel field events. + + +This measurability result does not change the deterministic proof or its constants. It supplies the missing event-theoretic foundation required to insert 1_{A_r} into exact expectations. + + +11. EXACT PALM QUOTIENT + + +Let + + + W_r^{exact} + =|det H_M det H_S| + 1{H_M negative definite} + 1{H_S index one}, + + +and + + + Z_r^{exact}=E_{Q_{6,r}}[W_r^{exact}]. + + +W_r^{exact} is Borel because endpoint derivative evaluations and matrix determinants/eigenvalue inequalities are continuous finite-dimensional functions of f. + + +If 0