Creating place fields in place cells is crucial for forming spatially dependent episodic memories in the brain, with the hippocampus playing a central role. Specifically, we focus on CA1 pyramidal neurons, which receive inputs from the CA3 and entorhinal cortex (EC). No widely accepted network-based theory for place field formation exists, but many models offer mechanisms for place-field firing. One type of hypotheses proposes that it results from summated output of spatially tuned cells, like grid cells or boundary vector cells. Another suggests CA1 place fields are replicate of CA3 place fields. A study showed that CA1 place fields can be learned from random initial weights between CA3 and CA1 through Hebbian-like, unsupervised spike-timing-dependent plasticity (STDP) with synaptic weight competition. However, the physiological relevance of the global competition rule and the arbitrary nature of the resultant place selectivity pose significant theoretical challenges, such as how learned structures in CA3 recurrents get successfully transferred to CA1 for further processing. Our new model, based on previous studies of LTP triggered by Ca2+ plateau potentials in CA1 dendrites—a form of one-shot learning with seconds-long plasticity kernels—incorporates new mechanistic details from brain slice experiments from our lab. Different from the original slice work, which required intracellular blockade of dendritic potassium channels, our experiments successfully induced BTSP (behavioral time-scale synaptic plasticity) with temporally structured, multiple-pathway inputs from EC and CA3. We also found physiological evidence supporting the “third factor” nature of BTSP, over various possible neuromodulator inputs. Our goal was to provide comprehensive simulations specifying how one-shot plasticity with long plasticity kernels as well as supervision produces place fields in a CA3-CA1 network. The computational properties of the details of learning mechanism led to distinct principles for episodic memory encoding in the network. For instance, our electrophysiological characterization enabled realistic implementation of supervision for multi-pathway-based plateau learning rules. The preference of plateau-induced LTP over frequency-ramping dynamics of CA3 would suggest a role of neural trajectories in episodic representations. Linking electrophysiology and circuit functions, our efforts aim to elucidate memory-processing computations via the lens of synaptic algorithms.