Human-in-the-Loop
A system design where human feedback or intervention is integrated into the model training, evaluation, or decision-making process to ensure safety.
Think of It Like This
Like a self-driving car that handles highway cruising but alerts a human driver to take the wheel in chaotic city traffic.
Human-in-the-Loop (HITL) ensures models do not operate entirely autonomously when confidence is low or risks are high. It is commonly used in data labeling, model fine-tuning (like RLHF), and high-stakes inference environments. This paradigm dramatically improves system reliability and aligns outputs with human values.