Overview of ‘Keynotes & Tutorials’ - CCN 2025

Value of the Information & Strength of the Argument

The video provides valuable insights into current research on distributed working memory mechanisms, presenting a compelling argument that working memory is not solely dependent on prefrontal cortex attractor dynamics. The speaker supports his claims with references to experimental data (e.g., Fuster’s recordings, receptor density gradients) and computational modeling. The argumentation is logical and well-structured, moving from classical theories to modern challenges and then to his own model’s predictions. The value lies in its synthesis of recent literature and the presentation of a novel modeling framework that could shift perspectives on working memory.

Scientific Rigor, Source Quality, Title Accuracy

The speaker demonstrates scientific rigor by referencing specific studies (e.g., Fuster, Mongillo, Stokes, Wang) and using state-of-the-art anatomical datasets (e.g., Kennedy lab connectivity, Gelstone spine counts). He also mentions a recent paper by Clansman and Prois Walsh and corroborating autoradiography data. The sources are appropriate and credible. The title accurately reflects the content, as it is an overview of the keynote and tutorial sessions. The video is a conference introduction, so it does not delve into full methodological details, but it provides sufficient context and references for further exploration.

The title accurately reflects the content: it is an overview of the keynote and tutorial sessions at CCN 2025.

8/10

The video is a conference introduction to keynotes and tutorials, featuring a presentation by a computational neuroscientist. It provides a high-level overview of recent research on distributed working memory models, referencing specific studies and data. The speaker is an expert in the field, and the content is scientifically grounded, though it is a summary rather than a detailed exposition.

Markers derived by PSI from the transcript: the creator did not define chapters.

• 00:00Introduction to the keynote and tutorials session, welcoming attendees.
• 02:00Speaker introduces himself and the topic: digital brain models of working memory.
• 05:00Discussion of classical working memory models based on attractor dynamics and prefrontal cortex.
• 10:00Presentation of recent challenges to classical models, including dynamic encoding and silent working memory.
• 15:00Evidence for distributed working memory activity across cortical areas.
• 20:00Introduction of the macaque full-brain model and its ability to reproduce distributed persistent activity.
• 25:00Explanation of the counterstream inhibitory bias and its importance in the model.
• 30:00Discussion of a study predicting decreasing NMDA-to-AMPA ratio along the cortical hierarchy.
• 35:00Extension of the model to the human brain using T1/T2 ratio and receptor densities.
• 40:00Summary and invitation to the tutorial for hands-on experience.

• CCN 2025 Conference Website — Mentioned as the conference where the talk was given.

• Fuster, J. M. (1973). Unit activity in prefrontal cortex during delayed-response performance. — Classic study showing sustained activity in prefrontal cortex during working memory delay.
• Mongillo, G., Barak, O., & Tsodyks, M. (2008). Synaptic theory of working memory. — Proposes a synaptic mechanism for working memory without persistent spiking.

• Classical attractor models of working memory — The speaker acknowledges that classical models focusing on prefrontal cortex attractor dynamics may not fully explain distributed working memory activity.