Human-in-the-loop

Human-in-the-loop or HITL is defined as a model that requires human interaction. HITL is associated with modeling and simulation (M&S) in the live, virtual, and constructive taxonomy. HITL models may conform to human factors requirements as in the case of a mockup. In this type of simulation a human is always part of the simulation and consequently influences the outcome in such a way that is difficult if not impossible to reproduce exactly. HITL also readily allows for the identification of problems and requirements that may not be easily identified by other means of simulation.

Human-in-the-loop - Wikipedia Jump to content From Wikipedia, the free encyclopedia Software user interface Part of a series on Machine learning and data mining Paradigms Supervised learning Unsupervised learning Semi-supervised learning Self-supervised learning Reinforcement learning Meta-learning Online learning Batch learning Curriculum learning Rule-based learning Neuro-symbolic AI Neuromorphic engineering Quantum machine learning Problems Classification Generative modeling Regression Clustering Dimensionality reduction Density estimation Anomaly detection Data cleaning AutoML Association

related reading

• Interaction Models: A Scalable Approach to Human-AI Collaboration - Thinking Machines Labthinkingmachines.ai
• Simulators — LessWronglesswrong.com
• In (highly contingent!) defense of interpretability-in-the-loop ML training — AI Alignment Forumalignmentforum.org
• Simulation - Wikipediaen.wikipedia.org
• Today, we are releasing a research preview of our user model, along with a set of evaluations designed to measure how faithfully user models capture human behavior.persimmon.humansand.ai
• A Functional Taxonomy of World Models - Dr. Fei-Fei Lidrfeifei.substack.com
• RLHF: Reinforcement Learning from Human Feedbackhuyenchip.com
• Previewing the Model Hardware Standardanthropic.com
• Andon Labsandonlabs.com
• The Coming Loop | Armin Ronacher's Thoughts and Writingslucumr.pocoo.org
• Language Models, World Models, and Human Model-Buildinglingo.csail.mit.edu
• Andrej Karpathy — AGI is still a decade awaydwarkesh.com