Self-Evolving Agent Harness
Paper formalizes persistent self-evolution of agent harnesses with fixed weights and validation bounds.
The paper formalizes harness self-evolution: feedback-driven revisions to an agent's prompts, memory, workflows, tools, code, and plugins while model weights remain fixed. Accepted revisions persist across runs, and the framework separates update targets from proposal and admission. Under stated assumptions, the authors derive an expected generalization bound, a retention bound tied to historical regression budgets, and a high-probability admission rule using fresh validation data. A 66-page discussion covers theory, algorithms, evaluation, applications, and impacts on scientific reliability, physical safety, and accountability, with code at the Harness-RSI repository.
- Harness revisions change prompts, memory, workflows, and code while weights stay fixed.
- Accepted changes persist across runs, with separate proposal and admission mechanisms.
- Analysis gives generalization, retention, and high-probability admission bounds.
- Authors argue continual validation and independent safeguards are required for deployment.
From the original article
1 min read at huggingface.co
Harness self-evolution revises an agent's software scaffolding, including prompts, memory, workflows, and executable components, while keeping model weights fixed. We formalize a feedback-driven process in which accepted revisions persist across runs. The framework separates update targets from proposal and admission mechanisms.
By Chuyang Wei, Haoxiang Guan, Minghang Zhu · huggingface.co
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.