SNT CoreX Dynamics

— Research  ·  Develop  ·  Evolve —

Est. 2025 · Antalya, Turkey

SNT CoreX Dynamics is an independent, pre-startup research initiative founded by Durhan Yazır, developing Spectral Nod Theory — a unified mathematical framework spanning Lie algebra, neural dynamics, and large language model inference optimisation. Built independently, published peer-reviewed.

4+Active Papers
G₇Core Algebra
φ★CoreX Spark
2025Founded

About the Organisation

Who We Are

SNT CoreX Dynamics is a research organisation advancing a unified theoretical framework across mathematics, neuroscience, and machine learning — built independently, published peer-reviewed.

Our Origin

SNT CoreX Dynamics is an independent, pre-startup deep-tech research initiative founded in 2025 by Durhan Yazır in Antalya, Turkey. It was established to develop and formalise Spectral Nod Theory — a program bridging pure mathematics with applied machine learning and biological neural dynamics. Operating without institutional affiliation, all research directions are selected based on mathematical necessity and scientific merit alone. The initiative is on a clear roadmap toward becoming a formally registered research organisation.

What We Build

Our work is organised around the SNT framework — seven canonical operators acting on structured state spaces, anchored by the seven-operator Lie algebra G₇ and the CoreX Spark constant φ★ = (1+√5)/2. From this algebraic foundation, we derive results in operator theory, connectome dynamics, and KV-cache optimisation for large language models. Every result is computationally verified and reproducible.

How We Operate

All manuscripts target peer-reviewed journals across mathematics, computer science, and interdisciplinary science. Every paper includes complete formal proofs, SymPy-verified computations, and open experimental pipelines. Papers share operator definitions, algebraic constants, and benchmarking methodology — forming a coherent program rather than isolated contributions. Affiliation on all submissions: SNT CoreX Dynamics, Independent Research Initiative, Antalya, Turkey.

Our Commitment

SNT CoreX Dynamics maintains a strict standard: no results are presented without formal derivation or experimental validation. We do not speculate — we derive, simulate, verify, and publish. The CoreX Spark constant φ★, the G₇ algebra structure, and all LLM benchmarks are fully documented and independently replicable. As an independent initiative, transparency is not optional — it is the foundation of credibility.

Road to Organisation

SNT CoreX Dynamics currently operates as an independent, pre-startup research initiative — a deliberate choice that preserves full intellectual autonomy during the foundational phase of the SNT program. The trajectory is clear: independent researcher today, research collective or registered institute tomorrow. As peer-reviewed publications accumulate and the SNT framework matures, the initiative will transition toward formal organisation — whether as a deep-tech research startup, a registered research institute, or a collaborative entity. This transition will be driven by the strength of the science, not administrative timelines. In the interim, all work is conducted with the rigour, transparency, and accountability expected of any established research body.

[∂]

Research

Continuously exploring new theoretical horizons with rigorous mathematical foundations.

[</>]

Develop

Transforming theoretical results into verified, reproducible computational systems.

[⟳]

Evolve

Iterating across domains — mathematics, biology, AI — within a unified framework.

[◈]

Autonomy

Full research independence, unconstrained by institutional agenda or external mandate.

[≡]

Rigour

Scientific honesty, transparency, and ethical standards in every result we publish.

[→]

Impact

Contributing meaningfully to human knowledge through peer-reviewed, reproducible science.

Mission

"To build a unified mathematical theory of structured dynamics — and to apply it, rigorously, across every domain it touches: from the algebraic structure of Lie algebras to the inference efficiency of large language models."

Publications & Manuscripts

Research

All manuscripts target peer-reviewed journals with complete proofs, verified computations, and reproducible experimental pipelines. Papers share the SNT operator framework across domains.

SNT-CORE · Lie Algebra · Pure Mathematics ◉ Late-Stage Preparation

Structural Obstruction in a Seven-Operator Lie Algebra: Derivation Rigidity and the CoreX Spark Spectral Constant

This paper establishes G₄(φ) ⊕ su(2) as the unique admissible closed Lie algebra structure over a seven-operator system. Three core theorems are formally proven: (T1) Jacobi identities close uniquely — the key commutator [F,N] = −i·tan(φ)·Cyc is derived rather than postulated; (T2) the spectral obstruction tensor is non-vanishing, rigorously ruling out semidirect product decomposition; (T3) the outer derivation algebra is one-dimensional and uniquely realised at the CoreX Spark constant φ★ = (1+√5)/2, identified as the Perron-Frobenius spectral radius of the structure matrix. All results are verified via SymPy with reproducible code in the appendix.

Target JournalAIMS Mathematics
DomainLie Algebra · Operator Theory
Key Constantφ★ = (1+√5)/2
ShrikeSNT · KV-Cache · Machine Learning ◉ Active — Pipeline Complete

Reversible KV Cache Compression via Memory Inertia: A Stability-Oriented Approach to Transformer Inference

ShrikeSNT introduces a stability-first KV-cache compression architecture for autoregressive transformer inference. The method applies SNT's Memory Inertia operator with a frozen-spark threshold anchored at the CoreX constant φ, routing key-value pairs through a Termop → Subspace → Loss pipeline. Evaluated on LongBench-v2 across Qwen2-7B and Qwen2.5-14B, SNT achieves 24.3–27% task accuracy at 2.33× faster decoding compared to StreamingLLM, maintaining 7/7 ablation consistency. Memory mass metric: 0.443 ± 0.013.

Target JournalSN Computer Science
ModelsQwen2-7B · Qwen2.5-14B
Key Result2.33× speedup vs StreamingLLM
BladeRunnerSNT · KV-Cache · Machine Learning ◉ Active

Drift-Aware Adaptive KV Cache Pruning for Efficient Transformer Inference

BladeRunnerSNT is the performance-oriented companion to ShrikeSNT, introducing drift-aware pruning at 35% full-KV retention. The system targets adaptive keep ratios via token-level entropy signals, query-aware Value boosting through the SNT Nexter operator, and multi-layer attention fusion. Evaluated across Qwen2-7B, Mistral-7B, and Qwen2.5-14B in BF16. Mistral-7B achieves +10 percentage points over full-KV baseline on targeted subtasks.

Target JournalSN Computer Science
ModelsQwen2-7B · Mistral-7B · Qwen2.5-14B
Key Result+10pp over full-KV (Mistral-7B)
SNT-LIFE · Connectomics · Neural Dynamics ◉ Active

Seven-Operator Learning and Memory Framework Validated on C. elegans Connectome Data

The SNT-LIFE series applies the seven-operator formalism to biological neural circuit dynamics. Validation against Cook 2019 and Kato 2015 C. elegans connectome datasets achieves trajectory correlation r = 0.986 at 86× parameter reduction relative to Wilson-Cowan models, with Lie algebraic closure verified at 4.82×10⁻¹⁵. Extended work covers predator-prey survival dynamics with a formally defined Survival Viability Condition (SVC), and N=100 population-level coevolutionary simulations producing r = −0.983, Cohen's d = 4.06 versus random baseline.

Target JournalsNeuroscience of Consciousness · Artificial Life (MIT)
Key Resultr = 0.986 · 86× param reduction

Research Program

Projects

Each project is a named thread within the SNT CoreX framework — combining formal mathematical theory, computational implementation, and empirical validation across domains.

CoreX SparkFoundation

Lie Algebra · Pure Mathematics

The algebraic foundation of the entire SNT program. Establishes the seven-operator Lie algebra G₇ = G₄(φ) ⊕ su(2) and derives the CoreX Spark constant φ★ = (1+√5)/2 as its unique spectral invariant. Derivation rigidity proven: the outer derivation space is exactly one-dimensional, with no semidirect product decomposition possible.

ShrikeSNTActive

Machine Learning · LLM Inference

Stability-oriented KV-cache compression for autoregressive transformer inference. Implements the Memory Inertia operator with a frozen-spark threshold at φ★. LongBench-v2 validated on Qwen2-7B and Qwen2.5-14B; 2.33× decoding speedup over StreamingLLM with higher task accuracy. Mass metric 0.443 ± 0.013, 7/7 ablation consistency.

BladeRunnerSNTActive

Machine Learning · LLM Inference

Performance-oriented companion to ShrikeSNT. Drift-aware adaptive KV pruning at 35% full-KV retention with entropy-driven keep ratios and query-aware Value boosting via the Nexter operator. Multi-model evaluation across Qwen2-7B, Mistral-7B, and Qwen2.5-14B in BF16.

SNT-LIFEActive

Connectomics · Neural Dynamics · Ecology

Application of the seven-operator formalism to biological systems. C. elegans connectome validation (r = 0.986, 86× parameter reduction), predator-prey survival dynamics with formal SVC, and N=100 coevolutionary population simulations (r = −0.983, d = 4.06).

TarrasqueSNTIn Reserve

Machine Learning · Next-Generation KV

Next-generation hybrid KV scoring combining SNT operator signals with segment-level retention guarantees and a global KV memory bank. Designed to unify ShrikeSNT's stability model with BladeRunnerSNT's adaptive performance into a single coherent inference system.

SNT GravityEarlier Work

Theoretical Physics · Quantum Systems

Spacetime emergence from Planck-scale nod networks with four operators on joint qubit-nod Hilbert spaces. Decoherence derivations via replica trick and Born-Markov approximation; NV-center T₂ scaling predictions as relative scaling laws.

The Researcher

About

SNT CoreX Dynamics is a pre-startup independent research initiative led by Durhan Yazır, working at the intersection of pure mathematics, computational neuroscience, and machine learning engineering.

Durhan Yazır
Founder · Lead Researcher
LocationAntalya, Turkey
AffiliationSNT CoreX Dynamics — Independent Research Initiative
Founded2025
Active Papers4 manuscripts
Target VenuesAIMS Mathematics · SN Computer Science · Artificial Life · Neuroscience of Consciousness
ORCID0009-0007-4304-2066
Contact[email protected]

Background

Durhan Yazır is an independent researcher working across pure mathematics, computational neuroscience, and machine learning engineering. In 2025 he founded SNT CoreX Dynamics — an independent, pre-startup research initiative through which he develops and publishes Spectral Nod Theory, a program spanning multiple scientific domains simultaneously. His research is registered under ORCID 0009-0007-4304-2066.

Without institutional affiliation, his work is self-directed and fully accountable: problems are selected on mathematical merit, results are computationally verified before submission, and every manuscript targets peer-reviewed publication. His approach is characterised by algebraic first-principles thinking — deriving results that connect apparently unrelated fields, from the structure theory of Lie algebras to the inference efficiency of billion-parameter language models.

Research Philosophy

Independent research carries a higher burden of proof than institutionally affiliated work. SNT CoreX Dynamics embraces this: every claim is formally derived or experimentally verified before submission, every computation is reproducible, and every manuscript is written to the standard of the most rigorous peer review.

The SNT framework is deliberately cross-domain. The same seven-operator algebra that governs C. elegans neural trajectories also defines the compression architecture of ShrikeSNT's KV-cache system. Spectral Nod Theory exists precisely to find the structural invariants that persist across scales and systems.

Research Domains

Lie Algebra Operator Theory Spectral Theory Neural Dynamics Connectomics Computational Theory LLM Inference Adaptive Inference Systems KV-Cache Optimisation Quantum Error Correction Fault-Tolerant Architectures Coevolutionary Systems Theoretical Physics

Get in Touch

Contact

For research correspondence, collaboration inquiries, or academic communication, reach out directly.

Direct Contact

All research correspondence is handled directly by the founder. Response time is typically within 48 hours.

[email protected]
OrganisationSNT CoreX Dynamics
LocationAntalya, Turkey
Founded2025
FocusSpectral Nod Theory — Mathematics, Neural Dynamics, LLM Systems
InquiriesResearch collaboration · Peer review · Academic correspondence
Send a Message