Performance of model-free algorithms on a 7-item pairwise serial learning task, presenting all 21 pairs during training and then reversing the order of the stimuli during testing

(A) Mean performance of Q-learning on non-terminal pairs throughout training and testing, sorted by symbolic distance. (B). Mean performance of Q-learning for each of the 21 pairs at transfer, sorted by symbolic distance. Critical pairs are shaded in gray. (C) Mean performance of Q-learning for each of the 21 pairs at the end of training. (D) to (F) As above, but reporting simulated performance of the Value Transfer Model (VTM). (G) to (I) As above, but reporting the simulated performance of RL-REMERGE