Search papers, labs, and topics across Lattice.
This paper introduces Semantic Lenia, a novel framework that redefines Large Language Model (LLM) inference as a continuous dynamical system rather than a static optimization task. By implementing a non-linear homeostatic feedback loop that balances semantic attraction with syntactic repulsion, the authors reveal the emergence of "Autonomous Semantic Solitons," which are stable structures that prevent repetitive outputs. Their extensive parameter sweeps identify a critical "Habitable Ridge" that allows for generative processes to operate at the edge of chaos, leading to significant cognitive advancements without structural failure.
Autonomous Semantic Solitons emerge from a new dynamical framework, enabling LLMs to generate diverse outputs while avoiding stagnation.
We introduce Semantic Lenia, an artificial life framework that transforms Large Language Model (LLM) inference from a static optimization problem into a continuous dynamical system within the macroscopic logit space. By establishing a non-linear homeostatic feedback loop to dynamically balance semantic attraction and syntactic repulsion, we demonstrate the emergence of"Autonomous Semantic Solitons"-- macroscopic dissipative structures that avoid repetitive crystallization. Our exhaustive parameter sweeps map a critical"Habitable Ridge"where applied steering forces perfectly balance the model's intrinsic syntactic inertia. This approach successfully maintains generative trajectories at the edge of chaos, triggering profound abductive leaps without structural collapse and establishing a physical scaling law for machine cognition.
