Experience-Handoff Reinforcement Learning for Recursive Self-Improvement
cs.LG, cs.CL
Submitted: 2026-11-12
Updated: 2026-11-12
Code: https://github.com/karpathy/autoresearch
Terminology
Sources
- GEPA: Reflective Prompt Evolution Can Outperform Reinforcement Learning
- A Survey of Self-Evolving Agents: What, When, How, and Where to Evolve on the Path to Artificial Super Intelligence
- Program Synthesis with Large Language Models
- Reflect, Retry, Reward: Self-Improving LLMs via Reinforcement Learning
- Large Language Monkeys: Scaling Inference Compute with Repeated Sampling
- Training-Free Group Relative Policy Optimization
- Composer 2 Technical Report
- Group-in-Group Policy Optimization for LLM Agent Training
- GLM-4.5: Agentic, Reasoning, and Coding (ARC) Foundation Models
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- SWE-bench: Can Language Models Resolve Real-World GitHub Issues?
- DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelines
- Kimi K2: Open Agentic Intelligence
- Meta-Harness: End-to-End Optimization of Model Harnesses
- CompactionRL: Reinforcement Learning with Context Compaction for Long-Horizon Agents
- Agentic Harness Engineering: Observability-Driven Automatic Evolution of Coding-Agent Harnesses
- Understanding R1-Zero-Like Training: A Critical Perspective
- Scaling LLM Multi-turn RL with End-to-end Summarization-based Context Management
- AlphaEvolve: A coding agent for scientific and algorithmic discovery
- Qwen3 Technical Report
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis