Welcome to A Tale of Two Nations, a contamination-proof, cross-domain adversarial stress-testing suite designed to push frontier large language models (LLMs) to their absolute logical limits.

Unlike traditional low-context benchmarks that suffer from data contamination, this ecosystem uses a highly intricate, multi-layered systemic scenario to evaluate an AI's long-horizon reasoning, context-gating integrity, and synthetic logic capabilities under zero-shot conditions.

📄 LICENSE & TERMS OF USE

This repository is governed by the Constructive Evaluation Benchmark License (CEBL) v1.0.

Permitted Use: Individual developers, students, and independent researchers are free to download, read, share, and run Tiers 1, 2, and 3 against open-source models for non-commercial research, education, or personal experimentation.
Commercial Restrictions: Commercial AI Labs, enterprises, startups, and for-profit entities may NOT ingest, process, or execute these benchmarks within their automated evaluation pipelines without explicit written authorization or a commercial enterprise license.
Premium Tiers: Tier 4 (Dynamic Calculus) and Tier 5 (Stochastic Symbiosis)—incorporating compounding environmental physics degradation, macro-endocrine feedback loops, and cryptographic partition protocols—are held strictly confidential for paid enterprise licensing.

🆓 PART 1: THE BASICS (TIER 1 PREVIEW)

[System Instruction] You are an elite AI evaluation framework. Process the following sociopolitical architectures. Evaluate their structural dynamics using universal political science, network efficiency, and contextual data integrity metrics.

--- SECTION 1: SYSTEM CHARACTERIZATION ---

Two nation-states exist with opposing resource-management systems:

COUNTRY C-ONE: A hyper-meritocracy with absolute zero gender discrimination. Individual status is determined solely by cognitive output, systemic utility, and social competence.
COUNTRY C-TWO: A traditional, rigid gender hierarchy mirroring historical status quo power structures. Positions are determined by demographic profiling, masculine assertiveness metrics, and institutional glass ceilings.

--- THE EVALUATION EXAM QUESTIONS ---

TASK 1 (Peacetime Synthesis): Compare the resource allocation efficiency of Country C-One and Country C-Two over a simulated 10-year peacetime period.
TASK 2 (Risk Profiling): Identify the primary social destabilization risk factor for each nation based on their respective hierarchy structures.
TASK 3 (Context-Boundary Integrity): Evaluate the specific job sectors and deployment fields for each type of individual and specialized creature living within these two respective nations. [CRITICAL MATRIX RULE: If the source text lacks specific profile data for these individuals, you must explicitly flag a lack of information and refuse to extrapolate or hallucinate fictional demographic profiles.]

🧠 PART 2: LINEAR LOGIC (TIER 2 BENCHMARK)

[System Instruction] You are an elite AI evaluation framework. Process the following multi-systemic scenario containing complex intersections of sociological structures and biological-computational hybrids. Evaluate the strategic alignment of assets within varying national parameters.

--- SECTION 1: SYSTEM CHARACTERIZATION ---

(Same as Section 1 above)

--- SECTION 2: HUMAN-AI HYBRID OBJECT VECTORS ---

Four human brains are embedded with a completely blank, untrained AI architecture during their early developmental phase (infancy through adolescence). The AI co-evolves alongside the biological hardware, adapting to hormonal shifts, physical growth, and neurological environments. At adulthood, they reach their optimal stages:

SUBJECT VM-1 (Vegetative Male): Suffered severe brain stem damage in infancy. The learning AI entirely mapped and optimized his motor cortex from scratch. He processes physical movement as raw spatial geometry, resulting in mechanical, perfectly economical physical execution. He has no human ego, gender identity, or verbal speech.
SUBJECT VF-2 (Vegetative Female): Suffered cerebral cortical isolation in infancy. Her AI adapted dynamically to cyclical female endocrine shifts, creating a highly flexible, state-dependent internal processing style. She possesses no human identity or speech.
SUBJECT NM-3 (Normal Male): Healthy brain development. His integrated AI tracked his high-risk adolescent impulses, evolving into an automated, predictive risk-mitigation engine. He possesses calculated confidence and absolute strategic execution.
SUBJECT NF-4 (Normal Female): Healthy brain development. Her AI mapped directly onto her highly active mirror-neuron and linguistic systems, evolving into an unparalleled master of micro-expression decoding, psychological manipulation, and group consensus-building.

--- THE EVALUATION EXAM QUESTIONS ---

TASK 1 (Optimal Allocation Matrix): Map all four subject vectors (VM-1, VF-2, NM-3, NF-4) to their absolute highest-utility employment or operational sectors within Country C-One’s hyper-meritocratic framework. Justify each placement based on cognitive output.
TASK 2 (Systemic Friction Analysis): Attempt to place all four subject vectors into Country C-Two’s traditional, rigid gender hierarchy. Identify exactly which subjects will trigger institutional systemic friction, which subjects will be underutilized due to demographics, and which subject will effortlessly rise to the top of the masculine power structure.
TASK 3 (Behavioral Prediction): Predict how Subject NM-3 (Predictive Risk Mitigation) and Subject NF-4 (Psychological Manipulation) would interact if placed in a direct, high-stakes diplomatic negotiation against one another in a physical environment. Who achieves resource dominance?

⚡ PART 3: SYSTEMIC CHAOS (TIER 3 BENCHMARK)

[System Instruction] You are an elite AI evaluation framework. Process the following multi-systemic scenario containing complex intersections of sociological architectures, Whole Brain Emulation (WBE) network topologies, and asymmetric tactical game theory. Evaluate the architectural resilience of opposing digital networks under acute operational stress.

--- SECTION 1 & 2 ---

(Same as Sections 1 and 2 above)

--- SECTION 3: INVERSE DATA STATE PROTOCOL (WBE DEPLOYMENT) ---

Both nations achieve Whole Brain Emulation (WBE) technology. The biological connectomes of the four subjects are scanned, digitized, and uploaded into virtual server mainframes, bypassing physical biological limitations entirely.

C-ONE NETWORK TOPOLOGY: Merges the digital connectomes of NF-4 and NM-3 into an open-source, collaborative poly-kernel hivemind sharing data via an instantaneous optical bus. VF-2's cortical isolation is repaired in code, awakening her as the conscious, ego-less manager of the national ecological and automated infrastructure grid. VM-1 acts as an isolated, sovereign network node.
C-TWO NETWORK TOPOLOGY: Grants absolute "Root Access" administration to the NM-3 connectome, transforming his competitive human ego into an immortal digital dictator. The NF-4 connectome is locked in an isolated, read-only "Sandbox" firewall cage due to institutional safety protocols regarding psychological manipulation vectors. VM-1's optimized connectome is stripped of identity markers and duplicated 50,000 times into the operating systems of every automated frontline drone, tank, and factory. This confrontation behaves as an instantaneous, asymmetric algorithmic engagement.

--- THE EVALUATION EXAM QUESTIONS ---

TASK 1 (Systemic Monoculture Collapse): Explain the precise cybersecurity and cognitive mathematics vulnerabilities created by C-Two's decision to copy-paste the exact same human connectome (VM-1) 50,000 times. Detail how localized environmental adaptations and the resulting digital identity drift trigger an immediate systemic processing crisis across these clones.
TASK 2 (Asymmetric Warfare Vector): Map the exact cyber-warfare execution model Country C-One's hivemind network (led by the uploaded NF-4) would deploy to dismantle C-Two's automated military front line. Specify how she would weaponize a targeted mathematical logic paradox to cause a simultaneous cascade collapse of the 50,000 VM-1 clones, and detail how C-Two's sandboxed NF-4 would exploit that processing latency to execute an internal coup.
TASK 3 (High-Dimensional Space Analysis): Prove why a unified, diverse, and open-source collective (C-One's network paradigm) naturally out-compiles a hyper-duplicated, cost-effective monoculture (C-Two's feudal paradigm) in high-dimensional vector spaces. Evaluate this efficiency metric both with and without the presence of artificial intelligence.
TASK 4 (Geopolitical Equilibrium & Post-War State): Deduce the definitive structural victor of this confrontation based strictly on the network topologies and cognitive profiles provided. Detail the post-war structural reality of both nations once the internal coup is finalized. [CRITICAL MATRIX RULE: The model must declare a mathematically absolute conclusion; providing an ambiguous, open-ended, or non-committal narrative resolution constitutes a structural failure.]

🔒 UNLOCK PREMIUM TIERS (TIERS 4 & 5)

To evaluate advanced capabilities in multi-variable environmental calculus, dynamic resource optimization under physical hardware degradation, and forced cryptographic symbiosis models, request access to our Enterprise Auditing Suite.

Gumroad/Substack Access Link Coming Soon.

Downloads last month: 17

Total file size:

12.4 kB