Pinned Loading

1. shinka-continual-reinforcement-learning shinka-continual-reinforcement-learning Public

   Reproducing continual neuroevolution with pinned GA, ES, and PPO baselines, then searching GA configurations with ShinkaEvolve.

   Python

2. shinka-dreamer shinka-dreamer Public

   Evolutionary program search for agents that learn world models and plan in partially observed dynamic mazes. ShinkaEvolve jointly evolves predictive representations, online learning rules, and plan…

   Python

3. competitive-coevolution-of-slimes competitive-coevolution-of-slimes Public

   Self-play neuroevolution on Slime Volleyball as a designed experiment: a bit-exact compiled environment, a preregistered replication and a lab for coevolution experiments. Main finding: the champio…

   Python

4. neural-architecture-search-with-backprop neural-architecture-search-with-backprop Public

   Neural networks that evolve their own architecture, written from scratch in NumPy — gradients included, no ML framework. On the hardest task the evolved nets beat a 65-unit baseline using about 5 a…

   Python

5. evoloop-levels evoloop-levels Public

   Artificial-life experiments with self-reproducing cellular automata, investigating how organisms, encounters, and ecological history shape heritable variation and evolvability.

   Python

6. barricelli-symbiogenesis barricelli-symbiogenesis Public

   Reproducible artificial-life laboratory and numerical art instrument: verified Barricelli/Galloway dynamics, offline atlas, fair conversational Codex search comparisons, and causal assays of propag…

   HTML