Goal-oriented loops let models test and iterate on their own
Defining a workflow where the model can operate, test, and iterate is central to automation. A common pattern is a goal-oriented loop: plan → execute → observe/test → improve → execute again until the goal is achieved. The loop may proactively ask users for clarification on task specification or execution preferences.
Evidencia a favor
Harness Engineering for Self-Improvement
Extracto original
Defining a workflow in which the model can operate, test, and iterate is a key design for automation. Karpathy’s autoresearch repo ( https://github.com/karpathy/autoresearch ) is a clean example of how such a workflow can be constructed. A common workflow follows a goal-oriented loop of plan, execute, observe/test, improve, and execute again until the goal is achieved. The process may trigger proactive requests to users for clarity in task specification or execution preference.
Compartir informaciónVerificar esta afirmación