Tolman-Eichenbaum Machine

Learning occurs in the context of individual prior experience. Variability in learning is, therefore, due to both differences in learning capacity as well as differences in those prior experiences. To describe the effect of prior experience on learning and to identify how the brain transfers this prior experience for current learning, we measured the behavior of many rats across two tasks on a six-arm track. In the first, reward was delivered after any transition between arms. In the second, reward was only delivered at three arms (2-3-4). To receive reward, it was necessary to alternate betwe

The abstract representation of space has been extensively studied in the hippocampus and entorhinal cortex. Yet while a growing variety of theoretical models have been proposed to capture the rich neural and behavioral phenomenology associated with these regions, it remains difficult to compare these theories systematically against each other using the full range of empirical data. To address this gap, we built an open-source standardised software framework, named NeuralPlayground, to facilitate comparisons of hippocampus and entorhinal cortex models. This Python software package offers a repr

Recent theoretical work has proposed that hippocampus acts as a binding site for sensory stimuli and a task coordinate system, each represented in the entorhinal cortex. Since this coordinate system path-integrates, incoming sensory observations can be inferred rather than learned. Here, we show that (1) a similar mechanism can learn representations that predict actions rather than stimuli. This mechanism supports zero-shot action prediction – a significant departure from current theoretical frameworks for action-learning. (2) The learned representations resemble object-vector and landmark cel