Chunk #6 — RESULTS — Further behavioral evidence for model-based learning

Source

States versus rewards: dissociable neural prediction error signals underlying model-based and model-free reinforcement learning.

Embedded

yes

Text

(with a fixed learning rate), then we would not expect any influence of their particular sequence of trials in session 1 on their choices in session 2. Thus, in the case of no model-based learning in session 1, any sequence of trials (including the actually experienced sequence) should lead to the same quality of model fit to the choices, whereas in the case of state learning with the FORWARD model, we would expect a better model fit under the actual trial sequence compared to any other random sequence.