Risk evaluation and behaviour: defining appropriate frames of reference
Many biological, psychological and economic experiments have been designed where an organism or individual must choose between two options that have the same expected reward but differ in the variance of reward received. In this way, designed empirical approaches have been developed for evaluating risk preferences. Here, however, we show that if the experimental subject is inferring the reward distribution (to optimize some process), they will never agree in finite time that the expected rewards are equal. In turn, we argue that this makes discussions of risk preferences, and indeed the motivations of behaviour, not so simple or straightforward to interpret. We use this particular experiment to highlight the serious need to consider the frame of reference of the experimental subject in studies of behaviour.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
Defining Admissible Rewards for High Confidence Policy Evaluation
A key impediment to reinforcement learning (RL) in real applications with limited, batch data is defining a reward function that reflects what we implicitly know about reasonable behaviour for a task and allows for robus…
Off-policy evaluationReinforcement LearningReinforcement Learning (RL)Vocal Bursts Intensity PredictionBenchmarkCards: Large Language Model and Risk Reporting
Large language models (LLMs) offer powerful capabilities but also introduce significant risks. One way to mitigate these risks is through comprehensive pre-deployment evaluations using benchmarks designed to test for spe…
FairnessLanguage ModelingLanguage ModellingLarge Language Model+1Multiple Appropriate Facial Reaction Generation in Dyadic Interaction Settings: What, Why and How?
According to the Stimulus Organism Response (SOR) theory, all human behavioral reactions are stimulated by context, where people will process the received stimulus and produce an appropriate reaction. This implies that i…
Defining and Evaluating Decision and Composite Risk in Language Models Applied to Natural Language Inference
Despite their impressive performance, large language models (LLMs) such as ChatGPT are known to pose important risks. One such set of risks arises from misplaced confidence, whether over-confidence or under-confidence, t…
Language ModellingNatural Language InferenceCollective Invasion: When does domain curvature matter?
Real-world cellular invasion processes often take place in curved geometries. Such problems are frequently simplified in models to neglect the curved geometry in favour of computational simplicity, yet doing so risks ina…