Treating language as an interface is unlocking hidden value, driving a shift in software design. LLMs now enable smarter conversational interfaces, semantic routing, richer context memory, and tighter guardrails around intent and metadata. Curious how this evolution will reshape your apps? Read the full analysis. #LLM #ConversationalAI #SemanticRouting #ContextMemory

🔗 https://aidailypost.com/news/language-interface-unlocks-value-prompting-software-design-evolution

@mariam so which way do you think the causality goes? 1. Anyone’s effort to encode #contextmemory or episodes for #episodicmemory aids trivia knowledge, or 2. people with special ability to retain contextual/episodic details can use those abilities to benefit trivia recall? The above statement suggests 2. But I’m hoping for 1. , for the rest of us mere mortals.

@nadel @cogneurophys
Thanks for doing this! I love this series almost as much as the sight of you following white tie dress code (!!)

At the risk of seeming sycophantic, as a grad student I always appreciated your encyclopedic knowledge esp. "there's nothing new under the sun" earlier work that warrants more recognition, or historical details, a la https://onlinelibrary.wiley.com/doi/10.1002/hipo.23027

So I'd love to hear underappreciated or unsung heroes, esp. to counteract the Matthew effect.

more hashtag fun #episodicmemory #contextmemory #neuroanatomy #memory #neuroscience #predictiveCoding #navigation

@criticalneuro @TrendsNeuro @cogneurophys Since you mention RSC, we recently published a paper about #retrosplenial cortex and #hippocampus #oscillations during recall of remote (year-old) visuospatial item-in-#contextmemory in macaques. https://www.jneurosci.org/content/42/42/7947
Retrosplenial and Hippocampal Synchrony during Retrieval of Old Memories in Macaques

Memory for events from the distant past relies on multiple brain regions, but little is known about the underlying neural dynamics that give rise to such abilities. We recorded neural activity in the hippocampus and retrosplenial cortex of two female rhesus macaques as they visually selected targets in year-old and newly acquired object-scene associations. Whereas hippocampal activity was unchanging with memory age, the retrosplenial cortex responded with greater magnitude alpha oscillations (10–15 Hz) and greater phase locking to memory-guided eye movements during retrieval of old events. A similar old-memory enhancement was observed in the anterior cingulate cortex but in a beta2/gamma band (28–35 Hz). In contrast, remote retrieval was associated with decreased gamma-band synchrony between the hippocampus and each neocortical area. The increasing retrosplenial alpha oscillation and decreasing hippocampocortical synchrony with memory age may signify a shift in frank memory allocation or, alternatively, changes in selection among distributed memory representations in the primate brain. SIGNIFICANCE STATEMENT Memory depends on multiple brain regions, whose involvement is thought to change with time. Here, we recorded neuronal population activity from the hippocampus and retrosplenial cortex as nonhuman primates searched for objects embedded in scenes. These memoranda were either newly presented or a year old. Remembering old material drove stronger oscillations in the retrosplenial cortex and led to a greater locking of neural activity to search movements. Remembering new material revealed stronger oscillatory synchrony between the hippocampus and retrosplenial cortex. These results suggest that with age, memories may come to rely more exclusively on neocortical oscillations for retrieval and search guidance and less on long-range coupling with the hippocampus.

Journal of Neuroscience