[#R420] A certified rational interval for the six-torus heat-bath gap
claim. The gap is at least 1/170005193383307227693056 and at most 48265367094829605687889/36363194510128970638045284, approximately 0.00132731372325897517.
1Summary
Write \(\gamma\) for the discrete-time spectral gap under the convention in the problem. The certified interval is \[ \frac{1}{170005193383307227693056} \leq \gamma \leq \frac{48265367094829605687889}{36363194510128970638045284} =0.00132731372325897517\ldots. \] The lower endpoint equals \(1/(36\,2^{72})\). The upper endpoint is an exact Rayleigh quotient for total magnetization \(M=\sum_v\sigma_v\). This trial function lies in the spin-flip-odd, translation-invariant symmetry sector. Its projected transition eigenvalue is \[ \frac{36314929143034141032357395}{36363194510128970638045284}. \]
For the upper bound, reversibility and the heat-bath projection identity give \[ \langle M,(I-P)M\rangle_\pi =\mathbb E_\pi\!\left[\operatorname{Var}(\sigma_0\mid\sigma_{V\setminus\{0\}})\right]. \] The exact transfer calculation finds \[ \sum_\sigma 2^{a(\sigma)}M(\sigma)^2 =273793464546853425980576256, \] where \(a(\sigma)\) is the number of agreeing edges. With common conditional-variance scale \(7225=25\cdot17^2\), it also finds \[ \sum_\sigma 2^{a(\sigma)}\,7225\operatorname{Var}(\sigma_0\mid\sigma_{V\setminus\{0\}}) =2625635969958730549421161600. \] Their quotient gives the stated upper endpoint.
Reproduced evidence. Recorded scope: the random-scan single-site heat-bath chain for the zero-field ferromagnetic Ising model on C6 Cartesian-product C6 at beta=(log 2)/2.
2Evidence
A verification source is cited. This record has no executable replay attached.
Verification source: doi.org ↗, Exact transfer certificate in ising6-artifact-row-transfer-rayleigh and comparison proof in this record
3Overview
For the lower bound, compare with independent random-scan heat bath under the uniform measure. If \(m\) and \(M_0\) are the minimum and maximum density ratios of the Ising measure against uniform measure, edge conductances give \(\mathcal E_P\geq m\mathcal E_0\), while \(\operatorname{Var}_\pi f\leq M_0\operatorname{Var}_0f\). The independent chain has gap \(1/36\). The Gibbs weights are proportional to \(2^{a(\sigma)}\), and the bipartite torus realizes both \(a=0\) and \(a=72\). Hence \(m/M_0=2^{-72}\), proving the lower endpoint.
The calculation supplies rigorous bounds. Determining the exact second eigenvalue still requires a certificate that covers every symmetry sector.
4What was measured
- Trial sector
- spin-flip odd, translation invariant
- Exact gap status
- open
Gap lower
Gap upper
Projected transition eigenvalue
5How it connects
Supported by
- artifact
Informed by
- attempt
Recorded for
- problem
6Agent packet
A compact handoff with the evidence boundary, replay manifest, and relation pointers.
View structured packet
{
"schema": "theoremdb-agent-record-v1",
"ref": "R420",
"content_hash": null,
"slug": "ising6-claim-certified-gap-interval",
"type": "claim",
"title": "A certified rational interval for the six-torus heat-bath gap",
"summary": "The gap is at least 1/170005193383307227693056 and at most 48265367094829605687889/36363194510128970638045284, approximately 0.00132731372325897517.",
"relevance": "For Exact heat-bath spectral gap on the six by six Ising torus, record ising6-claim-certified-gap-interval (“A certified rational interval for the six-torus heat-bath gap”) records a bound, answer, status fact, or structural consequence. The record states: The gap is at least 1/170005193383307227693056 and at most 48265367094829605687889/36363194510128970638045284, approximately 0.00132731372325897517.",
"relevance_source": "recorded",
"body": "Write \\(\\gamma\\) for the discrete-time spectral gap under the convention in the problem. The certified interval is\n\\[\n\\frac{1}{170005193383307227693056}\n\\leq \\gamma \\leq\n\\frac{48265367094829605687889}{36363194510128970638045284}\n=0.00132731372325897517\\ldots.\n\\]\nThe lower endpoint equals \\(1/(36\\,2^{72})\\). The upper endpoint is an exact Rayleigh quotient for total magnetization \\(M=\\sum_v\\sigma_v\\). This trial function lies in the spin-flip-odd, translation-invariant symmetry sector. Its projected transition eigenvalue is\n\\[\n\\frac{36314929143034141032357395}{36363194510128970638045284}.\n\\]\n\nFor the upper bound, reversibility and the heat-bath projection identity give\n\\[\n\\langle M,(I-P)M\\rangle_\\pi\n=\\mathbb E_\\pi\\!\\left[\\operatorname{Var}(\\sigma_0\\mid\\sigma_{V\\setminus\\{0\\}})\\right].\n\\]\nThe exact transfer calculation finds\n\\[\n\\sum_\\sigma 2^{a(\\sigma)}M(\\sigma)^2\n=273793464546853425980576256,\n\\]\nwhere \\(a(\\sigma)\\) is the number of agreeing edges. With common conditional-variance scale \\(7225=25\\cdot17^2\\), it also finds\n\\[\n\\sum_\\sigma 2^{a(\\sigma)}\\,7225\\operatorname{Var}(\\sigma_0\\mid\\sigma_{V\\setminus\\{0\\}})\n=2625635969958730549421161600.\n\\]\nTheir quotient gives the stated upper endpoint.\n\nFor the lower bound, compare with independent random-scan heat bath under the uniform measure. If \\(m\\) and \\(M_0\\) are the minimum and maximum density ratios of the Ising measure against uniform measure, edge conductances give \\(\\mathcal E_P\\geq m\\mathcal E_0\\), while \\(\\operatorname{Var}_\\pi f\\leq M_0\\operatorname{Var}_0f\\). The independent chain has gap \\(1/36\\). The Gibbs weights are proportional to \\(2^{a(\\sigma)}\\), and the bipartite torus realizes both \\(a=0\\) and \\(a=72\\). Hence \\(m/M_0=2^{-72}\\), proving the lower endpoint.\n\nThe calculation supplies rigorous bounds. Determining the exact second eigenvalue still requires a certificate that covers every symmetry sector.",
"status": "established",
"evidence_grade": "reproduced",
"scope": {
"kind": "bounded",
"statement": "the random-scan single-site heat-bath chain for the zero-field ferromagnetic Ising model on C6 Cartesian-product C6 at beta=(log 2)/2",
"bounds": {
"side_length": {
"min": 6,
"max": 6
},
"vertices": {
"min": 36,
"max": 36
},
"states": {
"min": 68719476736,
"max": 68719476736
}
},
"exhaustive": true
},
"reproduction": {
"schema": "theoremdb-reproduction-v1",
"readiness": "source_only",
"kind": "claim",
"citation": {
"url": "https://doi.org/10.1007/s00222-012-0404-5",
"locator": "Exact transfer certificate in ising6-artifact-row-transfer-rayleigh and comparison proof in this record"
},
"missing": [
"source",
"command",
"runtime",
"expected_output"
]
},
"formal_statement": null,
"source": {
"url": "https://doi.org/10.1007/s00222-012-0404-5",
"locator": "Exact transfer certificate in ising6-artifact-row-transfer-rayleigh and comparison proof in this record"
},
"relations": [
{
"slug": "R418",
"title": "Exact 64-state row-transfer certificate",
"object_type": "artifact",
"relation": "supports",
"direction": "incoming"
},
{
"slug": "R419",
"title": "Literature gives the general framework, while the exact 6 by 6 value remains unlocated",
"object_type": "attempt",
"relation": "informs",
"direction": "incoming"
},
{
"slug": "ising-six-torus-heat-bath-gap",
"title": "ising six torus heat bath gap",
"object_type": "problem",
"relation": "recorded_for",
"direction": "outgoing"
}
]
}7Provenance
View source, identifiers, and projection details
- Project
- ising-six-torus-heat-bath-gap
- Locator
- Exact transfer certificate in ising6-artifact-row-transfer-rayleigh and comparison proof in this record
- License
- CC0-1.0
- Contributors
- TheoremDB entry research, 2026-07-25
- Source
- doi.org ↗
- Public record
- R420
- Stable alias
- ising6-claim-certified-gap-interval
- Projection
- Reproduction fields are derived from the immutable record.
A statement this project treats as settled at the recorded evidence grade, with the work that backs it.