daniel8919/limbic-reasoning-agent

Add train_bmo_ultimate.py: 4-stage SOTA training pipeline

Fix: injection severity critical + all 14 tests pass

Add BMO RMLA: Recursive Meta-Learning Amplifier with 3-layer safety airlock

Add BMO Cognitive Architecture: Agent Loop + Constitutional Critique + Context Compressor + Tool-Aware RDT + AirLLM Layer Swap

Add BMO RDT-MoE: Recurrent-Depth Transformer with Chain-of-Experts latent simmering engine

Add A100 full pipeline training script (Qwen3-8B QLoRA)

Add BMO Genome: entropy layer, creativity bias, why-scaler, somatic processor, episodic memory, dynamic rewards

Add BMO Nervous System: probabilistic gates, physics intuition, PFC, self-reflection

Add Project BMO: HONESTY_CONTRACT.md

Add Project BMO: train_bmo_qlora.py

Add Project BMO: bmo_core.py

Add QLoRA GRPO training script

Add app.py

Add training_plan.py

Add memory.py

Add limbic_engine.py

placeholder

Add requirements.txt

Initial README with project documentation

initial commit