Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Ouroboros Thesis v15 — Knot Calculus for Governed Decision Receipts: R1/R2/R3 Audit-Reidemeister Conjecture, PAC-Bayes Governance Head, and Khipu-DAG Receipt Structure

Stephen P. Lutar

Published May 28, 2026• 0 citations
Publisher

Abstract

v14 lands Anatomy-Evolved-v1 on top of the v13 Λ-Invariant Stack. New material (Chapter 9): (1) Bohr complementarity engine — 12 frame-pairs, σA·σB ≥ ¼ floor, KS-18 contextuality witness (Cabello 1996), POVM verdict layer (Davies-Lewis 1970, Preskill 2015), QBist subjective credence (Fuchs-Schack 2013) — all wired into a11oy. (2) QKAN-FWP graft (arXiv:2605.06734) — DARUAN single-qubit data re-uploading activation + scalar-gated fast-weight programmer with Lean-proved Frobenius-norm boundedness (lutar-lean GatedBoundedness.lean) — wired into amaru as the new sequence-memory primitive (12.5k params vs. LSTM 167k baseline on Mackey-Glass + sine). (3) Dual-use coefficient registry (sentra) + 364-day fixed-cycle scheduler with 24-course operator rotation (terra) + 8-template verdict library (counsel) + ownership-opacity scorer (vessels, monotone in shellDepth). (4) Doctrine v6 ban-list runtime guard (carlota-jo). (5) Two new Lean theorems: gated_qkan_boundedness, two_witness_KS18_soundness. Series A acceptance: tsc --noEmit clean, 25/25 smoke tests pass, doctrine guard clean, one real bug fixed (POVM completeness Σ E_m = I — no bandaid).

DOI
10.5281/zenodo.20424995

Related papers

Ouroboros Thesis v15 — Knot Calculus for Governed Decision Receipts: R1/R2/R3 Audit-Reidemeister Conjecture, PAC-Bayes Governance Head, and Khipu-DAG Receipt Structure

Stephen P. Lutar

v14 lands Anatomy-Evolved-v1 on top of the v13 Λ-Invariant Stack. New material (Chapter 9): (1) Bohr complementarity engine — 12 frame-pairs, σA·σB ≥ ¼ floor, KS-18 contextuality witness (Cabello 1996), POVM verdict layer (Davies-Lewis 1970, Preskill 2015), QBist subjective credence (Fuchs-Schack 2013) — all wired into a11oy. (2) QKAN-FWP graft (arXiv:2605.06734) — DARUAN single-qubit data re-uploading activation + scalar-gated fast-weight programmer with Lean-proved Frobenius-norm boundedness (lutar-lean GatedBoundedness.lean) — wired into amaru as the new sequence-memory primitive (12.5k params vs. LSTM 167k baseline on Mackey-Glass + sine). (3) Dual-use coefficient registry (sentra) + 364-day fixed-cycle scheduler with 24-course operator rotation (terra) + 8-template verdict library (counsel) + ownership-opacity scorer (vessels, monotone in shellDepth). (4) Doctrine v6 ban-list runtime guard (carlota-jo). (5) Two new Lean theorems: gated_qkan_boundedness, two_witness_KS18_soundness. Series A acceptance: tsc --noEmit clean, 25/25 smoke tests pass, doctrine guard clean, one real bug fixed (POVM completeness Σ E_m = I — no bandaid).

•May 28, 2026•0 citations

Ouroboros Thesis v16 — Λ-Invariant Stack with Feynman Path-Integral Audit Closure, Gates Doctrine Codes, and Cross-Component Composite Invariant

Stephen P. Lutar

v14 lands Anatomy-Evolved-v1 on top of the v13 Λ-Invariant Stack. New material (Chapter 9): (1) Bohr complementarity engine — 12 frame-pairs, σA·σB ≥ ¼ floor, KS-18 contextuality witness (Cabello 1996), POVM verdict layer (Davies-Lewis 1970, Preskill 2015), QBist subjective credence (Fuchs-Schack 2013) — all wired into a11oy. (2) QKAN-FWP graft (arXiv:2605.06734) — DARUAN single-qubit data re-uploading activation + scalar-gated fast-weight programmer with Lean-proved Frobenius-norm boundedness (lutar-lean GatedBoundedness.lean) — wired into amaru as the new sequence-memory primitive (12.5k params vs. LSTM 167k baseline on Mackey-Glass + sine). (3) Dual-use coefficient registry (sentra) + 364-day fixed-cycle scheduler with 24-course operator rotation (terra) + 8-template verdict library (counsel) + ownership-opacity scorer (vessels, monotone in shellDepth). (4) Doctrine v6 ban-list runtime guard (carlota-jo). (5) Two new Lean theorems: gated_qkan_boundedness, two_witness_KS18_soundness. Series A acceptance: tsc --noEmit clean, 25/25 smoke tests pass, doctrine guard clean, one real bug fixed (POVM completeness Σ E_m = I — no bandaid).

•May 28, 2026•0 citations

Ouroboros Thesis v17 — Wheelerian Audit Closure, Shannon Doctrine Code, and the QEC-Evolved Agent Body

Stephen P. Lutar

v17 of the Ouroboros Thesis threads two foundational physics grafts and the full quantum-error-correction lineage through the v13 SZL Agent Body anatomy. Three named contributions are kernel-checked in Lean 4 (Mathlib v4.13.0) and mirrored by TypeScript runtime counterparts with passing test suites. §II Wheelerian audit closure — a bounded delayed-choice receipt-close operator (W = 1000 ticks); operationalises Wheeler (1978) for AI doctrine; 7 theorems / 10 kernel-checked examples; 18 runtime tests. §III Shannon doctrine code — the 4-level lattice {Bot, L1, L2, Top} admits a Shannon-optimal 2-bit prefix code with Kraft sum equal to 1 at equality; receipt-channel rate bound r ≤ B/2 for an audit-closure bit-rate budget B; 7 theorems / 12 examples; 18 runtime tests. §IV QEC-evolved anatomy — wraps the v13 Heart/Brain/Blood/Immune organs with a defensive cortex threading Hamming distance/weight at the substrate, Shor [[9,1,3]] receipt replication in the blood organ (Amaru), the CSS classical-to-stabilizer bridge in the brain organ, Kitaev surface-code vertex parity in the immune organ (Sentra), and the Shannon doctrine code as a new spine; 10 theorems / 17 examples; 70 runtime tests across the ecosystem. Totals: 24 Lean theorems / 39 kernel-checked examples / 0 sorry / 0 axiom; 108 runtime tests passing; doctrine v6 clean (0 violations across 20 scanned files). Companion Lean modules in szl-holdings/lutar-lean. Companion runtime in szl-holdings/platform/packages/doctrine-runtime. Ecosystem integrations in szl-holdings/a11oy, amaru, sentra, rosie. Predecessor DOIs in the concept chain: v14 10.5281/zenodo.20424992; v15 10.5281/zenodo.20424995; v16 10.5281/zenodo.20424996.

•May 28, 2026•0 citations

Ouroboros Thesis v17 — Wheelerian Audit Closure, Shannon Doctrine Code, and the QEC-Evolved Agent Body

Stephen P. Lutar

v17 of the Ouroboros Thesis threads two foundational physics grafts and the full quantum-error-correction lineage through the v13 SZL Agent Body anatomy. Three named contributions are kernel-checked in Lean 4 (Mathlib v4.13.0) and mirrored by TypeScript runtime counterparts with passing test suites. §II Wheelerian audit closure — bounded delayed-choice receipt-close operator (W = 1000 ticks); operationalises Wheeler (1978) for AI doctrine; 7 theorems / 10 kernel-checked examples; 18 runtime tests. §III Shannon doctrine code — the 4-level lattice {Bot, L1, L2, Top} admits a Shannon-optimal 2-bit prefix code with Kraft sum equal to 1 at equality; receipt-channel rate bound r ≤ B/2 for an audit-closure bit-rate budget B; 7 theorems / 12 examples; 18 runtime tests. §IV QEC-evolved anatomy — wraps the v13 Heart/Brain/Blood/Immune organs with a defensive cortex threading Hamming distance/weight at the substrate, Shor [[9,1,3]] receipt replication in the blood organ (Amaru), the CSS classical-to-stabilizer bridge in the brain organ, Kitaev surface-code vertex parity in the immune organ (Sentra), and the Shannon doctrine code as a new spine; 10 theorems / 17 examples; 70 runtime tests across the ecosystem. Totals: 24 Lean theorems / 39 kernel-checked examples / 0 sorry / 0 axiom; 108 runtime tests passing; doctrine v6 clean (0 violations across 20 scanned files). Predecessor DOIs in the concept chain: v14 10.5281/zenodo.20424992; v15 10.5281/zenodo.20424995; v16 10.5281/zenodo.20424996.

•May 28, 2026•0 citations

Gated QKAN-FWP: Scalable Quantum-inspired Sequence Learning

Kuo-Chung Peng, Samuel Yen-Chi Chen, Jiun-Cheng Jiang +16 more

Fast Weight Programmers (FWPs) encode temporal dependencies through dynamically updated parameters rather than recurrent hidden states. Quantum FWPs (QFWPs) extend this idea with variational quantum circuits (VQCs), but existing implementations rely on multi-qubit architectures that are difficult to scale on noisy intermediate-scale quantum (NISQ) devices and expensive to simulate classically. We propose gated QKAN-FWP, a fast-weight framework that integrates FWP with Quantum-inspired Kolmogorov-Arnold Network (QKAN) using single-qubit data re-uploading circuits as learnable nonlinear activation, known as DatA Re-Uploading ActivatioN (DARUAN). We further introduce a scalar-gated fast-weight update rule that stabilizes parameter evolution, supported by a theoretical analysis of its adaptive memory kernel, geometric boundedness, and parallelizable gradient paths. We evaluate the framework across time-series benchmarks, MiniGrid reinforcement learning, and highlight real-world solar cycle forecasting as our main practical result. In the long-horizon setting with 528-month input window and 132-month forecast horizon, our 12.5k-parameter model achieves lower scaled Mean Square Error (MSE), peak amplitude error, and peak timing error than a suite of classical recurrent baselines with up to 13x more parameters, including Long Short-Term Memory (LSTM) networks (25.9k-89.1k parameters), WaveNet-LSTM (167k), Vanilla recurrent neural network (11.5k), and a Modified Echo State Network (132k). To validate NISQ compatibility, we further deploy the trained fast programmer on IonQ and IBM Quantum processors, recovering forecasting accuracy within 0.1% relative MSE of the noiseless simulator at 1024 shots. These results position gated QKAN-FWP as a scalable, parameter-efficient, and NISQ-compatible approach to quantum-inspired sequence modeling.

cs.LGCross-listedarXiv:2605.06734•May 8, 2026•3 citations

Governed Type-Safe Triage: Qwen3.5-0.8B LoRA Distillation with a Refusal-Preserving Evaluation Gate

Stephen P. Lutar

Canonical consolidated thesis (supersedes drafts v6 and v7; the single DOI document). Presents THREE machine-checked Lean 4 theorems forming the verified core of the SZL Holdings governed-AI kernel, framed within a cross-disciplinary methodology: (1) Wave24 - under pure-dephasing Lindblad/GKSL dynamics the l1 coherence monotone C(t)=C0*exp(-gamma*t) is strictly antitone, decays to zero, and the Lambda-v5 closure floor is crossed at a unique time t*=(1/gamma)ln(qC0/Lambda_min); (2) Wave25 - a constructive Kleene iterate-supremum characterization of the Lambda-aggregator least fixed point given an explicit omega-continuity commute hypothesis; (3) Wave26 - the unconditional strengthening that derives that commute property from omega-Scott-continuity (full Kleene fixpoint theorem). All verified with Lean 4 v4.18.0 + Mathlib v4.18.0, no sorry, Lean-core axioms only, merged to main. HONEST NOVELTY: the contribution is the machine-checked formalization plus governance application; no new physics or pure mathematics is claimed (underlying results classical: Lindblad 1976, Baumgratz-Cramer-Plenio 2014, Kleene 1938, Tarski 1955). The cross-disciplinary lineage (Sherman Morgan constrained optimization, Stewart collision/EOS, BFT quorum, agentic routing) is adapted methodological structure, attributed to original authors, never reclaimed. Honest doctrine: locked-proven = exactly 8 {F1,F4,F7,F11,F12,F18,F19,F22} @ c7c0ba17; the three new theorems are EXPERIMENTAL/CI-green tier and NEVER join the locked 8; Lambda unconditional uniqueness stays Conjecture 1 (machine-checked FALSE; these are convergence results, not uniqueness); Khipu BFT = Conjecture 2; SLSA L1+L2 attested, L3 roadmap; trust never 100%. Jack Kruse = narrative only.

•Jun 6, 2026•0 citations
v14 lands Anatomy-Evolved-v1 on top of the v13 Λ-Invariant Stack. New material (Chapter 9): (1) Bohr complementarity engine — 12 frame-pairs, σA·σB ≥ ¼ floor, KS-18 contextuality witness (Cabello 1996), POVM verdict layer (Davies-Lewis 1970, Preskill 2015), QBist subjective credence (Fuchs-Schack 2013) — all wired into a11oy. (2) QKAN-FWP graft (arXiv:2605.06734) — DARUAN single-qubit data re-uploading activation + scalar-gated fast-weight programmer with Lean-proved Frobenius-norm boundedness (lutar-lean GatedBoundedness.lean) — wired into amaru as the new sequence-memory primitive (12.5k params vs. LSTM 167k baseline on Mackey-Glass + sine). (3) Dual-use coefficient registry (sentra) + 364-day fixed-cycle scheduler with 24-course operator rotation (terra) + 8-template verdict library (counsel) + ownership-opacity scorer (vessels, monotone in shellDepth). (4) Doctrine v6 ban-list runtime guard (carlota-jo). (5) Two new Lean theorems: gated_qkan_boundedness, two_witness_KS18_soundness. Series A acceptance: tsc --noEmit clean, 25/25 smoke tests pass, doctrine guard clean, one real bug fixed (POVM completeness Σ E_m = I — no bandaid).
v14 lands Anatomy-Evolved-v1 on top of the v13 Λ-Invariant Stack. New material (Chapter 9): (1) Bohr complementarity engine — 12 frame-pairs, σA·σB ≥ ¼ floor, KS-18 contextuality witness (Cabello 1996), POVM verdict layer (Davies-Lewis 1970, Preskill 2015), QBist subjective credence (Fuchs-Schack 2013) — all wired into a11oy. (2) QKAN-FWP graft (arXiv:2605.06734) — DARUAN single-qubit data re-uploading activation + scalar-gated fast-weight programmer with Lean-proved Frobenius-norm boundedness (lutar-lean GatedBoundedness.lean) — wired into amaru as the new sequence-memory primitive (12.5k params vs. LSTM 167k baseline on Mackey-Glass + sine). (3) Dual-use coefficient registry (sentra) + 364-day fixed-cycle scheduler with 24-course operator rotation (terra) + 8-template verdict library (counsel) + ownership-opacity scorer (vessels, monotone in shellDepth). (4) Doctrine v6 ban-list runtime guard (carlota-jo). (5) Two new Lean theorems: gated_qkan_boundedness, two_witness_KS18_soundness. Series A acceptance: tsc --noEmit clean, 25/25 smoke tests pass, doctrine guard clean, one real bug fixed (POVM completeness Σ E_m = I — no bandaid).
v14 lands Anatomy-Evolved-v1 on top of the v13 Λ-Invariant Stack. New material (Chapter 9): (1) Bohr complementarity engine — 12 frame-pairs, σA·σB ≥ ¼ floor, KS-18 contextuality witness (Cabello 1996), POVM verdict layer (Davies-Lewis 1970, Preskill 2015), QBist subjective credence (Fuchs-Schack 2013) — all wired into a11oy. (2) QKAN-FWP graft (arXiv:2605.06734) — DARUAN single-qubit data re-uploading activation + scalar-gated fast-weight programmer with Lean-proved Frobenius-norm boundedness (lutar-lean GatedBoundedness.lean) — wired into amaru as the new sequence-memory primitive (12.5k params vs. LSTM 167k baseline on Mackey-Glass + sine). (3) Dual-use coefficient registry (sentra) + 364-day fixed-cycle scheduler with 24-course operator rotation (terra) + 8-template verdict library (counsel) + ownership-opacity scorer (vessels, monotone in shellDepth). (4) Doctrine v6 ban-list runtime guard (carlota-jo). (5) Two new Lean theorems: gated_qkan_boundedness, two_witness_KS18_soundness. Series A acceptance: tsc --noEmit clean, 25/25 smoke tests pass, doctrine guard clean, one real bug fixed (POVM completeness Σ E_m = I — no bandaid).
v17 of the Ouroboros Thesis threads two foundational physics grafts and the full quantum-error-correction lineage through the v13 SZL Agent Body anatomy. Three named contributions are kernel-checked in Lean 4 (Mathlib v4.13.0) and mirrored by TypeScript runtime counterparts with passing test suites. §II Wheelerian audit closure — a bounded delayed-choice receipt-close operator (W = 1000 ticks); operationalises Wheeler (1978) for AI doctrine; 7 theorems / 10 kernel-checked examples; 18 runtime tests. §III Shannon doctrine code — the 4-level lattice {Bot, L1, L2, Top} admits a Shannon-optimal 2-bit prefix code with Kraft sum equal to 1 at equality; receipt-channel rate bound r ≤ B/2 for an audit-closure bit-rate budget B; 7 theorems / 12 examples; 18 runtime tests. §IV QEC-evolved anatomy — wraps the v13 Heart/Brain/Blood/Immune organs with a defensive cortex threading Hamming distance/weight at the substrate, Shor [[9,1,3]] receipt replication in the blood organ (Amaru), the CSS classical-to-stabilizer bridge in the brain organ, Kitaev surface-code vertex parity in the immune organ (Sentra), and the Shannon doctrine code as a new spine; 10 theorems / 17 examples; 70 runtime tests across the ecosystem. Totals: 24 Lean theorems / 39 kernel-checked examples / 0 sorry / 0 axiom; 108 runtime tests passing; doctrine v6 clean (0 violations across 20 scanned files). Companion Lean modules in szl-holdings/lutar-lean. Companion runtime in szl-holdings/platform/packages/doctrine-runtime. Ecosystem integrations in szl-holdings/a11oy, amaru, sentra, rosie. Predecessor DOIs in the concept chain: v14 10.5281/zenodo.20424992; v15 10.5281/zenodo.20424995; v16 10.5281/zenodo.20424996.
v17 of the Ouroboros Thesis threads two foundational physics grafts and the full quantum-error-correction lineage through the v13 SZL Agent Body anatomy. Three named contributions are kernel-checked in Lean 4 (Mathlib v4.13.0) and mirrored by TypeScript runtime counterparts with passing test suites. §II Wheelerian audit closure — bounded delayed-choice receipt-close operator (W = 1000 ticks); operationalises Wheeler (1978) for AI doctrine; 7 theorems / 10 kernel-checked examples; 18 runtime tests. §III Shannon doctrine code — the 4-level lattice {Bot, L1, L2, Top} admits a Shannon-optimal 2-bit prefix code with Kraft sum equal to 1 at equality; receipt-channel rate bound r ≤ B/2 for an audit-closure bit-rate budget B; 7 theorems / 12 examples; 18 runtime tests. §IV QEC-evolved anatomy — wraps the v13 Heart/Brain/Blood/Immune organs with a defensive cortex threading Hamming distance/weight at the substrate, Shor [[9,1,3]] receipt replication in the blood organ (Amaru), the CSS classical-to-stabilizer bridge in the brain organ, Kitaev surface-code vertex parity in the immune organ (Sentra), and the Shannon doctrine code as a new spine; 10 theorems / 17 examples; 70 runtime tests across the ecosystem. Totals: 24 Lean theorems / 39 kernel-checked examples / 0 sorry / 0 axiom; 108 runtime tests passing; doctrine v6 clean (0 violations across 20 scanned files). Predecessor DOIs in the concept chain: v14 10.5281/zenodo.20424992; v15 10.5281/zenodo.20424995; v16 10.5281/zenodo.20424996.
Fast Weight Programmers (FWPs) encode temporal dependencies through dynamically updated parameters rather than recurrent hidden states. Quantum FWPs (QFWPs) extend this idea with variational quantum circuits (VQCs), but existing implementations rely on multi-qubit architectures that are difficult to scale on noisy intermediate-scale quantum (NISQ) devices and expensive to simulate classically. We propose gated QKAN-FWP, a fast-weight framework that integrates FWP with Quantum-inspired Kolmogorov-Arnold Network (QKAN) using single-qubit data re-uploading circuits as learnable nonlinear activation, known as DatA Re-Uploading ActivatioN (DARUAN). We further introduce a scalar-gated fast-weight update rule that stabilizes parameter evolution, supported by a theoretical analysis of its adaptive memory kernel, geometric boundedness, and parallelizable gradient paths. We evaluate the framework across time-series benchmarks, MiniGrid reinforcement learning, and highlight real-world solar cycle forecasting as our main practical result. In the long-horizon setting with 528-month input window and 132-month forecast horizon, our 12.5k-parameter model achieves lower scaled Mean Square Error (MSE), peak amplitude error, and peak timing error than a suite of classical recurrent baselines with up to 13x more parameters, including Long Short-Term Memory (LSTM) networks (25.9k-89.1k parameters), WaveNet-LSTM (167k), Vanilla recurrent neural network (11.5k), and a Modified Echo State Network (132k). To validate NISQ compatibility, we further deploy the trained fast programmer on IonQ and IBM Quantum processors, recovering forecasting accuracy within 0.1% relative MSE of the noiseless simulator at 1024 shots. These results position gated QKAN-FWP as a scalable, parameter-efficient, and NISQ-compatible approach to quantum-inspired sequence modeling.
Canonical consolidated thesis (supersedes drafts v6 and v7; the single DOI document). Presents THREE machine-checked Lean 4 theorems forming the verified core of the SZL Holdings governed-AI kernel, framed within a cross-disciplinary methodology: (1) Wave24 - under pure-dephasing Lindblad/GKSL dynamics the l1 coherence monotone C(t)=C0*exp(-gamma*t) is strictly antitone, decays to zero, and the Lambda-v5 closure floor is crossed at a unique time t*=(1/gamma)ln(qC0/Lambda_min); (2) Wave25 - a constructive Kleene iterate-supremum characterization of the Lambda-aggregator least fixed point given an explicit omega-continuity commute hypothesis; (3) Wave26 - the unconditional strengthening that derives that commute property from omega-Scott-continuity (full Kleene fixpoint theorem). All verified with Lean 4 v4.18.0 + Mathlib v4.18.0, no sorry, Lean-core axioms only, merged to main. HONEST NOVELTY: the contribution is the machine-checked formalization plus governance application; no new physics or pure mathematics is claimed (underlying results classical: Lindblad 1976, Baumgratz-Cramer-Plenio 2014, Kleene 1938, Tarski 1955). The cross-disciplinary lineage (Sherman Morgan constrained optimization, Stewart collision/EOS, BFT quorum, agentic routing) is adapted methodological structure, attributed to original authors, never reclaimed. Honest doctrine: locked-proven = exactly 8 {F1,F4,F7,F11,F12,F18,F19,F22} @ c7c0ba17; the three new theorems are EXPERIMENTAL/CI-green tier and NEVER join the locked 8; Lambda unconditional uniqueness stays Conjecture 1 (machine-checked FALSE; these are convergence results, not uniqueness); Khipu BFT = Conjecture 2; SLSA L1+L2 attested, L3 roadmap; trust never 100%. Jack Kruse = narrative only.