Can a Language Model Improve Its Own Mathematical Reasoning? (Weak RSI) 2026 - present ·3 mins Recursive Self-Improvement LLM Training Mathematics LoRA Local AI Web Agents
