self-evolution

Google Cloud AI Research introduces SkillOS, a reinforcement learning framework enabling LLM-based agents to self-evolve by curating reusable skills from past experiences.

This paper introduces SEIF, a self-evolving reinforcement learning framework that enhances LLM instruction-following capabilities through iterative difficulty adaptation and co-training of instructor and follower components.

FlashEvolve is a framework that accelerates LLM-based agent self-evolution by using asynchronous stage orchestration and artifact version tracking. It significantly improves throughput and token efficiency compared to synchronous execution methods.

Claude agents have added a new 'Dreaming' feature that enables self-optimization by reviewing historical conversations and extracting patterns. Together with multi-agent parallel orchestration and quality assessment, this marks the transition of AI agents into a self-evolution stage.

SafeHarbor is a novel framework for LLM agent safety that uses hierarchical memory and self-evolution to balance safety and utility, achieving state-of-the-art performance on benign and malicious tasks.

RemoteZero is a framework that eliminates the need for human-annotated box supervision in geospatial reasoning by leveraging the semantic verification capabilities of multimodal large language models (MLLMs) to enable self-evolving localization from unlabeled remote sensing data.

This paper proposes a method to train LLM agents with intrinsic meta-evolution capabilities, enabling spontaneous self-improvement without external rewards at inference time. Applied to Qwen3-30B and Seed-OSS-36B, the approach yields a 20% performance boost on web navigation benchmarks, with a 14B model outperforming Gemini-2.5-Flash.

MiniMaxAI releases MiniMax-M2.7, an open-weight model featuring self-evolution capabilities, advanced agent team support, and strong performance on software engineering benchmarks (56.22% on SWE-Pro, 66.6% medal rate on MLE Bench Lite), with notable applications in production incident recovery and professional work tasks.

EvoScientist is an adaptive multi-agent framework for end-to-end scientific discovery that continuously improves through persistent memory modules, comprising three specialized agents for idea generation, experiment execution, and knowledge distillation. It outperforms 7 state-of-the-art systems in scientific idea generation and improves code execution success rates through multi-agent evolution.