Reinforcement learning models of risky choice and the promotion of risk-taking by losses disguised as wins in rats.

摘要: Risky decisions are inherently characterized by the potential to receive gains or incur losses, and these outcomes have distinct effects on subsequent decision-making. One important factor is that individuals engage in loss-chasing, which reception of a loss followed relatively increased risk-taking. Unfortunately, mechanisms loss-chasing poorly understood, despite importance for understanding pathological choice behavior. The goal present experiment was illuminate governing individual differences risky-choice behaviors. Rats chose between low-uncertainty outcome always delivered variable amount reward high-uncertainty probabilistically reward. Loss-processing were assessed context losses disguised as wins (LDWs), presented along with gain-related stimuli. LDWs been suggested interfere adaptive decision-making humans thus potentially increase loss-making. Here, rats riskier, they made more choices outcome. A series nonlinear models fit rats' data elucidate possible psychological best account prone showing stay bias following compared not LDWs. These results collectively suggest acquire conditioned reinforcement properties encourage continued risk-taking previous high-risk decisions. (PsycINFO Database Record