HITL
Human in the Loop (HITL) integrates human judgment into the decision-making process of AI systems. Crucial for ensuring AI reliability and alignment with human values.
9 topics found for:
“human judgment”
Representativeness
Representativeness is a heuristic in decision-making where individuals judge the probability of an event based on how much it resembles a typical case. Crucial for understanding biases in human judgment and improving decision-making processes.
RLHF
Reinforcement Learning from Human Feedback (RLHF) is a machine learning technique that uses human input to guide the training of AI models. Essential for improving the alignment and performance of AI systems in real-world applications.
Regressive Bias
A cognitive bias where individuals overestimate the likelihood of extreme events regressing to the mean. Crucial for understanding decision-making and judgment under uncertainty.
Category Size Bias
A cognitive bias where people judge the likelihood of an event based on the size of its category rather than its actual probability. Crucial for designers to understand how category size influences user perception and decision-making processes.
Prospect Theory
A behavioral economic theory that describes how people choose between probabilistic alternatives that involve risk, where the probabilities of outcomes are known. Crucial for understanding decision-making under risk and designing systems that align with user behavior.
Anchoring
Anchoring (also known as Focalism) is a cognitive bias where individuals rely heavily on the first piece of information (the "anchor") when making decisions. Crucial for understanding and mitigating initial information's impact on user decision-making processes.
Egocentric Bias
A cognitive bias where people rely too heavily on their own perspective and experiences when making decisions. Important for designers to recognize and mitigate their own perspectives influencing design decisions.
Subadditivity Effect
A cognitive bias where the total probability assigned to a set of events is less than the sum of the probabilities assigned to each event individually. Important for understanding how users estimate probabilities and make decisions under uncertainty.