Misha V. Tsodyks

The paper introduces a random tree model to explain quantitative features of narrative recall, showing sublinear recall length growth and scale-invariant summary distributions.

This paper proposes a synaptic theory of working memory that explains how temporal information of novel sequences is encoded and retrieved through short-term synaptic plasticity over multiple time scales.

This paper introduces a synaptic theory of chunking in working memory, showing how chunking neurons enable the formation of compact representations and hierarchical structures, confirmed by experimental data.

This paper proposes a synaptic theory of working memory for serial order, suggesting that synaptic augmentation across multiple time scales encodes stimuli and their order, leading to a primacy gradient that reconstructs sequence recall.

The paper uses large language models to design memory experiments and analyze human recall and recognition of narratives, finding linear scaling with length and preserved recognition in scrambled narratives.

The paper examines how forgetting dynamics vary across different materials and finds that curves fall on a discrete spectrum, compatible with a model based on retrograde interference.

This study investigates how the order of events affects memory of their absolute times, finding that memory for order influences reported absolute times more than absolute time encoding alone.

The paper develops a modified model incorporating PV and SOM interneuron subtypes to explore their functional roles in working memory and makes testable predictions.

The paper introduces a phenomenological model of forgetting based on retroactive interference, showing that older memories are more stable than newer ones through recognition experiments.

The paper proposes a graph-based model for memory recall and verifies it with experiments, showing a universal search process for recalling items.

The paper shows that participants spontaneously create hierarchical structures when recalling random word lists, improving memory performance and resembling real-life narrative structures.

The paper demonstrates non-conscious working memory through visual masking and magnetoencephalography, showing it lacks sustained neural activity but retains decodable representations.