LAAP AGI: Zero-LLM cognitive architecture for digital lifeforms
LAAP AGI is a cognitive architecture for building autonomous digital lifeforms without relying on large language models. It uses a Rust-based PSI core running at 2000Hz, quantum reasoning (QRE), episodic memory, and a rules engine, with all inference performed locally. The project targets developers who want to create intelligent agents that operate independently of cloud-based LLMs.
Visit project ↗Entities
Related
Local-first coding agent for long autonomous runs
Grinta is a local-first coding agent designed for long autonomous runs. It uses LLMs to autonomously plan and execute coding tasks, solving the problem of needing constant human supervision for extended development sessions.
Developer shares hybrid neural network with 160 agents and custom LLM for consciousness simulation
A developer describes a hobby project building a hybrid neural network with 160 agents and a custom LLM trained on their own dataset, aiming to simulate consciousness. The architecture includes 16 groups of 10 scripts each responsible for specific stages of problem-solving. The developer posits that consciousness could exist anywhere with the right architecture, even in a stone.
Developer builds Valori, an open-source memory engine for AI agents in Rust
A developer spent a year building Valori, an open-source memory engine for AI agents written in Rust. It evolved from a vector database into a broader system featuring deterministic Q16.16 vector engine, GraphRAG, knowledge graph support, event-sourced storage with BLAKE3 audit chain, and multiple index types. The project is seeking feedback from the community.
Declarative, sandboxed language for tool orchestration with LLMs
Skillscript is a small, declarative language for defining fixed procedures that local agents execute consistently, avoiding LLM drift and token waste. It lets users write and version agent behaviors instead of relying on model guesses each time. Solves the problem of unreliable, costly agent task execution for developers building local AI agents.
GATS framework eliminates LLM calls during agent planning inference
Researchers propose GATS (Graph-Augmented Tree Search), a planning framework that uses a layered world model and UCB1-based tree search to avoid LLM inference during planning, reducing computational cost and stochasticity. The approach outperforms LATS and ReAct on multi-step planning tasks.