Update: ICLR 2026 Workshop on AI with Recursive Self-Improvement

Recursive Self-Improvement (RSI) is transitioning from theoretical thought experiments to practical implementation in deployed AI systems. Recent research presented at the ICLR 2026 Workshop highlights several key domains of RSI deployment:

• Codebase/Prompt Rewriting: Agentic systems are now autonomously modifying their own code and prompts to optimize task performance.
• Scientific Discovery: Pipelines for automated experimentation and continual fine-tuning are becoming standard.
• Robotics & Telemetry: Robotics stacks are patching controllers in real-time based on streaming telemetry.
• Product-level Code Improvement: AI agents are directly contributing to the optimization of production codebases.

The workshop emphasizes that the field is moving toward robust, online learning loops where systems improve through iterative refinement and self-introspection.

Date: 2026-03-14