Press Releases
Press Releases
BUDAPEST / — EXXOGEN today announced the hardware validation of its continuous-time geometric quantum model, executing real-world quantum verification on Rigetti’s 108-qubit Cepheus-1-108Q superconducting quantum processor via Azure Quantum. The benchmark confirms that quantum entanglement and the CHSH inequality violation arise directly as a physical consequence of EXXOGEN’s internal dynamics rather than an axiomatic postulate.
First-Principles Tsirelson Convergence: EXXOGEN’s deterministic dynamics yielded a theoretical CHSH parameter of 2.82842712474619, matching the fundamental quantum bound of 2.8284271247461903.
Physical QPU Entanglement Violation: Executed on the Rigetti Cepheus-1-108Q hardware, the mapped gate sequence achieved a measured S-value of 2.404, comfortably exceeding the classical limit (S ≤ 2.000) and experimentally proving non-local entanglement.
High-Fidelity Quantum Benchmarks: Demonstrated a 0.9966 Quantum Teleportation fidelity, a 1.0000 (100%) Error Correction success rate, and a Quantum Volume Heavy Output Probability of 0.8272 (significantly above the 2/3 threshold).
Invariant Conservation Under Noise: Continuous dynamical evaluation confirmed that radius fluctuations (R_A, R_B) and decoherence parameters (|R_A - R_B|) strictly preserve internal geometric invariants (I), naturally driving the Bell correlation curve E(a,b) into the exact quantum cosine distribution.
Zero-Knowledge Hardware Execution: Local geometric invariants were mapped to standard single-qubit rotation gate parameters (θ, φ), allowing remote cloud execution on the QPU while maintaining complete IP isolation of EXXOGEN's closed-form continuous equations.
Unlike heuristic models or unverified simulators, EXXOGEN bridges continuous-time geometric determinism with standard quantum gate execution, preserving full proprietary logic offline while producing verified physics on third-party quantum hardware.
Theoretical CHSH Bound: 2.82842712474619 (EXXOGEN Model) vs. 2.8284271247461903 (Quantum Boundary)
QPU Execution Target: Rigetti Cepheus-1-108Q (108 Superconducting Qubits)
CHSH Violation Outcome: Measured S = 2.404
2026.09.28
///
First-principles simulations of Tau aggregates (7PX3) reveal quantized energy wells and dynamic THz-frequency vibrations, offering fresh physical insights into neurodegenerative protein structures.
BUDAPEST / — EXXOGEN, a DeepTech molecular simulation engine provider, today announced new findings from its physical analysis of the Tau amyloid fibril (7PX3), a structure strongly linked to Alzheimer’s disease.
Using EXXOGEN’s proprietary, closed-form analytical operator, the engine evaluated the physical dynamics of the Tau backbone, identifying structural elasticity and collective vibrational behavior within the Terahertz (1.2–1.3 THz) frequency range.
Quantized Energy Wells: The structural backbone exhibits distinct energy minima at 2.25 Å and 3.25 Å.
Structural Elasticity: The bound backbone demonstrates a stable spring constant range of 0.3 – 0.8 N/m.
Terahertz Vibrational Peak: Internal atomic vibrations show a sharp collective peak at 1.2 – 1.3 THz.
Understanding the physical and mechanical properties of Tau aggregates offers a new dimension for Alzheimer's studies. By mapping these specific vibrational frequencies, EXXOGEN’s data provides a potential foundation for researchers to explore non-chemical, frequency-based diagnostic and therapeutic modalities alongside traditional pharmacology.
"Mapping the exact physical and thermodynamic behavior of macromolecular structures gives researchers new avenues to explore," said the Founder at EXXOGEN. "Identifying discrete vibrational frequencies in Tau fibrils provides a clear physical dataset that can support new research in biophysics and neurodegeneration."
2026.09.24
///
BUDAPEST / — EXXOGEN today published validation suite, releasing raw datasets and execution scripts that allow institutional partners to independently verify its deterministic molecular simulation infrastructure.
Baseline Benchmark (PDB: 1D66)
Sub-Ångström Precision: 0.000499 Å spatial accuracy across 1,363 evaluated interaction vectors.
Physical Energy Landscape: r = -0.9005 correlation between energy and distance.
Symmetry Preservation: Strict translational and rotational SE(3) invariance (0.000499 Å max error).
Thermodynamic Stability: Validated Boltzmann partition function with a 2.305x bound-to-thermal energy separation ratio.
Extreme Scale Validation (Titin: Q8WZ42)
1.1M+ Vector Scalability: 1,117,942 interaction vectors evaluated deterministically on the largest human protein dataset.
Unbroken Sub-Ångström Accuracy: 0.000514 Å spatial precision with zero numerical drift under extreme scale.
Macro-Scale Potential Surface: r = -0.8888 correlation maintained across more than one million vectors.
Consistent Physics: SE(3) symmetry preserved with a 2.312x bound-to-thermal energy separation ratio.
Unlike data-fitted machine learning models, EXXOGEN computes interactions directly from first-principles physics via a proprietary, closed-form analytical operator. The validation suite and datasets are available for independent verification.
Data & Script Access: Github
2026.09.21