Sofiya Zbaranska
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Sofiya Zbaranska has a total of 13 co-authors. Here are the top co-authors by number of shared publications:
- Sheena A. Josselyn
Program in Neurosciences and Mental Health, The Hospital for Sick Children, Toronto, Canada
2 shared publications - Aditya Rajeev
University of Toronto, Toronto, Canada
1 shared publication Alessandro Luchetti
Program in Neuroscience and Mental Health, Hospital for Sick Children
1 shared publicationAlexander D. Jacob
Program in Neuroscience and Mental Health, Hospital for Sick Children
1 shared publicationAndrew Mocle
Program in Neuroscience and Mental Health, Hospital for Sick Children
1 shared publicationAnnelies Hoorn
Program in Neuroscience and Mental Health, Hospital for Sick Children
1 shared publication
Latest Publications
- Open Access
- 12 May 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (20)
Temporally gated offline engram ensemble reverberation in the lateral amygdala is required for fear memory consolidation
Memories are encoded by sparsely distributed neuronal ensembles, or engrams, yet how newly allocated engram neurons are consolidated to support subsequent memory retrieval remains unclear. While offline reactivation has been extensively studied in hippocampal circuits, whether similar ensemble-level dynamics support consolidation of emotional memories in the amygdala is unknown. Here we combine longitudinal in vivo calcium imaging, temporally precise activity-dependent neuronal tagging, and optogenetic manipulations to examine postlearning dynamics of fear engram ensembles in the lateral amygdala (LA) of mice. We find that neurons allocated to a fear engram exhibit a transient, temporally gated increase in spontaneous activity and functional coordination immediately after learning, which decay within hours. Disrupting LA activity during this early posttraining window, but not at later time points, produces persistent amnesia. Selective inhibition of neurons active immediately after training is sufficient to impair memory, whereas inhibition of neurons active 12 h posttraining or neurons responsive to an unrelated sensory stimulus has no effect. Critically, memories disrupted by posttraining engram inhibition cannot be rescued by direct optogenetic reactivation of the engram ensemble, indicating a failure of engram consolidation rather than a retrieval deficit. We further show that engram ensemble activity is similarly required following memory retrieval, implicating a shared circuit-level mechanism for consolidation and reconsolidation. Together, these findings identify offline engram ensemble reverberation in the LA as a causal mechanism for transforming a transient, cellularly allocated trace into a stable, retrievable emotional memory.
- 12 May 2026
- Published by National Academy of Sciences in Proceedings of the National Academy of Sciences
- Vol. 123 (20)
Open AccessTemporally gated offline engram ensemble reverberation in the lateral amygdala is required for fear memory consolidation
Memories are encoded by sparsely distributed neuronal ensembles, or engrams, yet how newly allocated engram neurons are consolidated to support subsequent memory retrieval remains unclear. While offline reactivation has been extensively studied in hippocampal circuits, whether similar ensemble-level dynamics support consolidation of emotional memories in the amygdala is unknown. Here we combine longitudinal in vivo calcium imaging, temporally precise activity-dependent neuronal tagging, and optogenetic manipulations to examine postlearning dynamics of fear engram ensembles in the lateral amygdala (LA) of mice. We find that neurons allocated to a fear engram exhibit a transient, temporally gated increase in spontaneous activity and functional coordination immediately after learning, which decay within hours. Disrupting LA activity during this early posttraining window, but not at later time points, produces persistent amnesia. Selective inhibition of neurons active immediately after training is sufficient to impair memory, whereas inhibition of neurons active 12 h posttraining or neurons responsive to an unrelated sensory stimulus has no effect. Critically, memories disrupted by posttraining engram inhibition cannot be rescued by direct optogenetic reactivation of the engram ensemble, indicating a failure of engram consolidation rather than a retrieval deficit. We further show that engram ensemble activity is similarly required following memory retrieval, implicating a shared circuit-level mechanism for consolidation and reconsolidation. Together, these findings identify offline engram ensemble reverberation in the LA as a causal mechanism for transforming a transient, cellularly allocated trace into a stable, retrievable emotional memory.
- Open Access
A Brain-Inspired Framework for Memory Prioritization in Neural Networks Based on Valence
Improving long-term memory in artificial neural networks remains an open challenge. To address this, we developed a novel brain-inspired framework for memory prioritization based on the principle of emotional valence. Our framework includes: (i) a valence-weighted cross-entropy loss that scales the learning signal by the valence magnitude, analogous to neuromodulation; (ii) an amygdala-inspired module that learns high-dimensional valence embeddings; and (iii) a hippocampus-inspired module that integrates valence embeddings into the attention mechanism to modulate information retrieval. We demonstrated the generalization of our framework across spatial, episodic, and language-based memory tasks, consistently improving memory prioritization and long-term retention of high-salience information. In addition to improving long-term memory, we also showed that our framework can help mitigate the “lost-in-the-middle” problem in language modeling. More generally, this research provides further evidence of the potential of brain-inspired algorithms to advance the field of machine learning.
Open AccessA Brain-Inspired Framework for Memory Prioritization in Neural Networks Based on Valence
Improving long-term memory in artificial neural networks remains an open challenge. To address this, we developed a novel brain-inspired framework for memory prioritization based on the principle of emotional valence. Our framework includes: (i) a valence-weighted cross-entropy loss that scales the learning signal by the valence magnitude, analogous to neuromodulation; (ii) an amygdala-inspired module that learns high-dimensional valence embeddings; and (iii) a hippocampus-inspired module that integrates valence embeddings into the attention mechanism to modulate information retrieval. We demonstrated the generalization of our framework across spatial, episodic, and language-based memory tasks, consistently improving memory prioritization and long-term retention of high-salience information. In addition to improving long-term memory, we also showed that our framework can help mitigate the “lost-in-the-middle” problem in language modeling. More generally, this research provides further evidence of the potential of brain-inspired algorithms to advance the field of machine learning.