Vela - New Era of Intelligence
Beyond LLMs, Inspired by the Human Brain
Vela is a cutting-edge research project focused on building a new kind of foundational model—one that fundamentally diverges from current large language models, which rely heavily on pattern matching and pre-defined data distributions.
Instead of scaling statistical correlations, Vela dives deep into the mechanisms of the human brain—modeling thinking, understanding, and emotion with needle-level precision. It draws from neuroscience and cognitive science to create a system that learns continuously, reasons causally, and adapts dynamically—just like biological intelligence.
With Vela, we're taking the first true step toward Artificial General Intelligence (AGI)—where a model's performance isn't dictated by the amount of training data, but by its ability to understand, self-learn, and evolve.
Core Capabilities
Environment-Aware Intelligence
Understands its surroundings with exceptional precision and context sensitivity.
Causal Decision-Making
Makes intelligent, goal-driven decisions based on reasoning, not just patterns.
Human-Like Learning
Learns from experience in real time—no pretraining or labeled data needed.
Self-Evolving Architecture
Continuously adapts and rewires itself without external retraining.
Research Behind Vela
The Living Web: A Biologically-Inspired Multidimensional Neural Architecture for Artificial General Intelligence
The paper introduces the Living Web, a novel computational model designed to move artificial intelligence closer to Artificial General Intelligence (AGI). Inspired by the structure and functions of the human brain, this architecture departs from current transformer-based models by focusing on causal reasoning, dynamic adaptability, embodied cognition, and energy efficiency. It features a multidimensional neural topology where each node can act as both input and output, supporting dense connectivity and combinatorial processing. The Living Web aims to provide a scalable, biologically plausible alternative to current AI systems, capable of adapting to new environments, learning over time without forgetting, and reasoning about the world like a human.
Visit Paper