News & Views | Published:


More than two forms of Pavlovian prediction

Behavioural neuroscience and reinforcement learning theory distinguish between ‘model-free’ and ‘model-based’ computations that can guide behaviour. A recent study demonstrates that Pavlovian learning can give rise to behavioural responses that are not well accounted for by this existing dichotomy, suggesting that there may be greater complexity to the computations that underlie Pavlovian prediction.

Access optionsAccess options

Rent or Buy article

Get time limited or full article access on ReadCube.


All prices are NET prices.


  1. 1.

    Schultz, W., Dayan, P. & Montague, P. R. Science 275, 1593–1599 (1997).

  2. 2.

    Robinson, M. J. & Berridge, K. C. Curr. Biol. 23, 282–289 (2013).

  3. 3.

    Zhang, S., Mano, H., Ganesh, G., Robbins, T. & Seymour, B. Curr. Biol. 26, 52–58 (2016).

  4. 4.

    Dayan, P. & Berridge, K. C. Cogn. Affect. Behav. Neurosci. 14, 473–492 (2014).

  5. 5.

    Pool, E. R., Pauli, W. M., Kress, C. S. & O’Doherty, J. P. Nat. Hum. Behav. (2018).

  6. 6.

    Daw, N. D., Niv, Y. & Dayan, P. Nat. Neurosci. 8, 1704–1711 (2005).

Download references

Author information

Correspondence to Catherine A. Hartley.

Ethics declarations

Competing interests

The authors declare no competing interests.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark