Rats rapidly switch between retrospective and inferential value computations.

There are many ways to compute value. For instance, animals can compute value by learning from the past or by imagining future outcomes, but it is unclear if or how these computations interact. We used high-throughput training to collect statistically powerful datasets from 240 rats performing a temporal wagering task with hidden reward states. Rats adjusted how quickly they initiated trials and how long they waited for rewards across states, balancing effort and time costs against expected rewards. Statistical modeling revealed that animals computed the value of the environment differently when initiating trials versus when deciding how long to wait for rewards, even though these decisions were only seconds apart. This work reveals that sequential decisions use parallel value computations on single trials.