DavidVivancos/Neuraxon2LifeTS

Dataset Summary Neuraxon2LifeTS is an ultra-large-scale, high-fidelity time-series dataset tracking the micro-events, neural dynamics, and evolutionary lifecycle of artificial agents ("Nxers") in a simulated Game of Life environment.

Unlike snapshot-heavy or purely spatial datasets, Neuraxon2LifeTS provides a rigorous tick-by-tick log of neural network states. It records complex signaling, synaptic plasticity, neuromodulator fluctuations, and homeostatic adjustments across generations. With over 30 distinct relational tables and billions of recorded micro-events, this dataset is designed for advanced research into artificial life, dynamic neural networks, and biologically inspired reinforcement learning.

by David Vivancos https://www.vivancos.com/ & Dr. Jose Sanchez https://josesanchezgarcia.com/ for Qubic Science. https://qubic.org/

Paper Reference: Neuraxon 2.0 : A New Neural Growth & Computation Blueprint https://www.researchgate.net/publication/400868863_Neuraxon_V20_A_New_Neural_Growth_Computation_Blueprint

Dataset Details Version: 2.0.0

Last Updated: February 28, 2026

Total Agents (Nxers): 1,050,709

Data Structure The dataset is broken down into over 30 relational tables formatted as .csv files. The architecture is divided into Core Entities, Massive Temporal Snapshots, Event Logs, and Environment states.

1. Massive Temporal Snapshots (The Heavy Lifters) These tables represent the bulk of the dataset's volume, capturing the tick-by-tick state of the simulation:

synapse_snapshots.csv: 2.59 billion rows tracking weight (fast/slow/meta), integrity, and plasticity traces.

neuron_snapshots.csv: 480.7 million rows tracking membrane potential, energy levels, health, phase, and autoreceptors.

per_nxer_time_series.csv: 8.4 million rows detailing agent-level time series, including neuromodulator levels (Dopamine, Serotonin, Acetylcholine, Norepinephrine), network activity, and fitness.

2. Core Entities games.csv: Global simulation parameters, total ticks, and environmental thresholds.

nxers.csv: Agent characteristics, lineage, input/output states, and survival statistics.

neurons.csv / synapses.csv: Base structural data defining the ~69.8 million neurons and ~375.5 million synapses generated across all sessions.

3. Event Logs Specific events triggered during the simulation are meticulously logged for causal analysis:

   Event Table	Row Count	Description
   associativity_events.csv	153.0M	Hebbian learning and neighbor contributions.
   plasticity_events.csv	153.0M	Synaptic weight delta events.
   weight_evolution_events.csv	153.0M	Granular shifts in fast, slow, and meta weights.
   neuromodulator_events.csv	153.0M	Threshold crossings for DA, SER, ACH, and NE.
   threshold_modulation_events.csv	107.9M	Dynamic shifts in effective firing thresholds.
   homeostatic_events.csv	69.8M	Activity-driven homeostatic corrections.

4. Environment & Evolution world_grids.csv & foods.csv: Spatial and nutritional constraints of the environment.

hall_of_fame.csv: Records of top-performing agents, their survival times, and energy efficiency.

clan_history.csv: Evolutionary lineage, merging, and clan activity.

Targeted Research Areas & Mappings Neuraxon2LifeTS is specifically mapped to support complex, multi-disciplinary research themes. Key research verticals supported by the variables include:

Trinary Neuromodulation: Tracking fractions of excitatory/inhibitory signals alongside complex chemical modulators.

Synaptic Computation: Evaluating ionotropic vs. metabotropic contributions, synaptic integrity, and silent-to-active transitions.

Temporal Dynamics & Synchronization: Analyzing phase coherence, cross-frequency coupling (CFC), and phase-amplitude coupling (PAC) via oscillator tracking.

Chronoplasticity & Astrocytic Influence: Advanced mechanics utilizing eligibility traces (agmp_astrocyte), dynamic decay, and temporal weight shifts.

Inherited Metabolism & Circadian Rhythms: Correlation of daylight, temperature variance, resting states, and resting metabolism multipliers.

Usage Notes Due to the extreme size of the temporal tables (e.g., synapse_snapshots.csv is ~455 GB), it is highly recommended to stream this dataset or query specific tables based on your research domain rather than loading the entire dataset into memory.

Citation @dataset{Neuraxon2LifeTS, title={Neuraxon 2.0 Game of Life Time Series Research Dataset}, author={Vivancos, David and Sanchez, Jose}, year={2026}, publisher={Hugging Face}, version={2.0.0}, url={https://huggingface.co/datasets/DavidVivancos/Neuraxon2LifeTS} }

Authors & Curators David Vivancos / Artificiology Research - Qubic Science Dr. Jose Sanchez / UNIR - Qubic Science Contact: For questions or issues, please open a GitHub issue at https://github.com/DavidVivancos/Neuraxon.