SYNOMORPHBiological Nervous-System Simulation
How much useful behavior can emerge from biological connectivity and simplified neural dynamics?
ExploreRESEARCH
From neural structure to procedural intelligence.
We study how nervous systems turn perception and experience into behavior — and whether those principles can help machines turn repeated reasoning into reusable skills.
RESEARCH THEMES
Reasoning → Skill → Habit
Can successful LLM trajectories become persistent procedural intelligence?
ExploreBiological vs Artificial Networks
When does evolved neural topology provide computational advantages?
ExploreEmbodied Nervous Systems
How does the same neural architecture behave when attached to different bodies?
ExploreProcedural Memory
How can agents retain and reuse skills without repeatedly retraining a foundation model?
ExploreEfficient Agent Runtime
Can reusable skills reduce repeated model calls, tokens, latency, and cost while preserving task performance?
ExploreWE TEST IT
SYNOMORPH does not assume biological wiring is better. We intend to benchmark biological connectomes, artificial network baselines, learned policies, static and evolving skills, and LLM-only agents using task success, model calls, tokens, latency, compute, repeatability, adaptability, and failure recovery.
SYNOMORPH separates hypotheses, visualizations, experimental models, recorded experiments, and validated results. A nervous-system structure is not treated as a reconstructed mind, and intended efficiency benefits are not presented as measured performance.