A neural substrate of prediction and reward.
Level 5 - mechanism / opinion, no new human data
Narrative review and theoretical synthesis of animal neurophysiology and computational models without systematic review methodology
PubMed 9054347 · doi:10.1126/science.275.5306.1593
What was done
Review and theoretical synthesis connecting behavioral learning experiments, primate neurophysiological recordings of dopaminergic neurons, and quantitative models of adaptive optimizing control.
What was found
The abstract provides no quantitative data or statistical metrics. It describes qualitative findings that the fluctuating activity of primate dopaminergic neurons encodes changes or errors in predictions of future salient and rewarding events.
Why it matters
Established the foundational conceptual framework linking midbrain dopamine firing to computational reinforcement learning and reward prediction error signaling.
Limits
The abstract describes a broad conceptual synthesis rather than original experimental data or a systematic review; no sample sizes, specific primate numbers, or quantitative effect sizes are reported.
Cited by
- context Hardwired dopamine circuitry dictates that following a major success, a subsequent achievement will not feel satisfying unless it exceeds the magnitude of the previous one.