The Editor in Your Head: How the Amygdala Decides What You Will Remember

Awareness · Brain science · Awareness Series No. 06 · By Dr. Saurabh Kumar, MD (Psychiatry, AIIMS New Delhi) · Associate Professor & HOD, RDJM Medical College · Consultant Neuropsychiatrist, Healthy Minds India, Patna · Edition 1.0, October 2026

The amygdala and memory: a filter we did not know we had

On 27 October 2026 a team at Cedars-Sinai in Los Angeles reported in Nature Neuroscience what may be the clearest demonstration yet that the amygdala decides, in the very first moments of an experience, whether that experience will be stored.¹ The press release, carried by News-Medical, describes the amygdala as a filter that tags a small subset of what we see for long-term keeping and lets the rest pass through.²

I wrote in this series earlier this month about a different Nature Neuroscience paper, on the amygdala and cooperation.³ Two major papers in one month on the same walnut-sized structure, both recorded directly from inside the living human brain, tell you where neuroscience is heading. The amygdala is being rediscovered: not as an alarm bell, but as an editor.

For a clinic that sees memory complaints every day, from anxious students to families bringing a parent who has started repeating questions, this matters. Let me explain what was found, how, and what it may eventually mean.

How the amygdala memory study was done

As with the cooperation study, the data came from people with epilepsy who had fine electrodes placed deep in the brain to locate the source of their seizures. While they waited in hospital, 112 of them volunteered for a memory experiment. That is an unusually large number for research of this kind, assembled by a consortium funded by the US National Institutes of Health's BRAIN Initiative.²

The task was simple. Each person was shown a series of pictures they had never seen before: faces, animals, landscapes and other categories. Later the same day they were tested on which pictures they remembered. Throughout, the electrodes recorded the firing of individual neurons in the amygdala.

The question was whether anything in those neurons, at the moment a picture was first seen, could predict whether it would be remembered hours later.

What the study found about how memories are formed

Specialist neurons. The amygdala contains neurons that respond selectively to particular categories of image: some fire for faces, others for animals, others for scenes. This was already suspected. The new finding concerns what those neurons do for memory.

A prediction made in the first glance. The more strongly a category neuron fired when its preferred kind of image appeared, the more likely that image was to be recalled later. In other words, the amygdala's response during encoding (the first second or so of seeing) predicted later memory success. The researchers replicated this in two separate datasets, which is the standard that separates a finding from a fluke.^1,2

Not only for frightening things. The classical view held that the amygdala boosts memory only for emotional events: the accident, the insult, the exam result. This study found the effect for ordinary, neutral pictures. The amygdala appears to be tagging everyday experience, not just the dramatic kind.¹

Think of a newspaper editor reading the wires. Hundreds of stories arrive; a few get the front page. The hippocampus has long been understood as the printing press. This study says the amygdala is the editor who decides what gets printed at all.

Why it matters clinically. The authors, led by first author Jiayang Xiao with senior author Ueli Rutishauser, who directs the Center for Neural Science and Medicine at Cedars-Sinai, suggest that the memory failures seen in Alzheimer's disease and temporal lobe epilepsy may be partly failures of this amygdala filter, not only of the hippocampus. If so, the amygdala becomes a candidate target for future treatments using neuromodulation or cell-based therapies.²

Three things this memory research does not mean

1. 1It does not mean a treatment exists.

   The study identifies a mechanism and a possible target. No trial has stimulated the amygdala to improve memory in patients, and the amygdala lies deep in the brain, well beyond the reach of standard non-invasive stimulation. Anyone offering "amygdala therapy for memory" today is ahead of the science.

2. 2It does not mean the hippocampus is less important.

   The hippocampus remains the structure that binds an experience into a memory. The new work adds a gatekeeper in front of it; it does not remove the gate.

3. 3It does not yet cover sound, smell or touch.

   The experiment used pictures only. Whether the amygdala filters other kinds of experience the same way is, by the authors' own account, the next question.

What it may mean for memory disorders: dementia, epilepsy and anxiety

Here I move from what the paper shows to what a clinician reads into it, and I will mark the boundary clearly.

Dementia. The amygdala is damaged early in Alzheimer's disease, earlier than most people realise, alongside the hippocampus.⁴ Families often describe a stage, before the obvious forgetting, when a parent seems to stop registering new things: a visit happens but leaves no trace. That description fits a failing filter as well as a failing store. If future studies confirm the amygdala's role, it could help explain why some patients encode nothing new while still retrieving old memories well, and it would make the amygdala a reasonable place to look for intervention. This is plausible, not proven. (Our page on dementia care in Patna explains how memory is assessed today.)

Epilepsy. People with temporal lobe epilepsy, the most common form in adults, frequently complain of poor memory even when seizures are controlled. The amygdala sits in the temporal lobe and is often involved.⁴ The study offers a direct reason why. (See our page on epilepsy treatment.)

Depression and anxiety. The amygdala is over-active in anxiety and under-regulated in depression, and both conditions impair memory for everyday events while sharpening memory for threat. A filter biased towards danger would store exactly what anxious patients tell me they cannot stop remembering, and discard exactly what they tell me they forget. The paper did not study this, but it gives the observation a mechanism to be tested.

Brain stimulation. At Healthy Minds India we use rTMS, theta-burst stimulation and tDCS, all of which act on the brain's surface. The amygdala cannot be reached directly by these methods, but it is densely connected to prefrontal regions that can. One reasonable reading of the new paper is that part of what prefrontal stimulation achieves in depression may run through the amygdala downstream. That is a hypothesis for research, and I state it only as that.

How to improve memory: four everyday lessons from the amygdala study

Basic science rarely translates into advice, but this study does line up with things we already know about how people remember. For students, professionals and anyone worried about their memory:

Your questions

Frequently Asked Questions

• Isn't the amygdala the "fear centre"?

  That label is outdated. The amygdala evaluates what matters: socially, emotionally and, we now see, for memory. Fear is one of the things it flags, not the only one.

• Can I strengthen my amygdala to improve my memory?

  Not in any targeted way, and you should not try. What you can do is give it better material: focused attention, meaningful connections, adequate sleep, and treatment for anxiety or low mood if those are present.

• My mother has early dementia. Does this amygdala research change her treatment?

  Not today. Current treatment remains medication where indicated, structured routine, physical activity, social engagement and family support. The study may shape future therapies; it does not alter present ones.

• Was the study done on healthy people?

  On people with epilepsy, because that is the only ethical setting for recording single neurons in humans. The large sample and replication in two datasets make the finding robust, but confirmation in other groups will follow.

• Can rTMS or brain stimulation treat memory disorders by reaching the amygdala?

  Not directly; it is too deep. Stimulation of connected surface regions may influence it indirectly, and deep-brain approaches are a research possibility. No clinic should claim to stimulate the amygdala with a standard coil.

• Why do I remember embarrassing moments so vividly but forget everyday things?

  Partly because strong emotion drives the amygdala's filter hard, so those moments are tagged with high priority. The new finding is that the same filter also works, more quietly, on the ordinary moments in between.

About this article

Written and medically reviewed by

Dr. Saurabh Kumar

• MD (Psychiatry) AIIMS New Delhi
• Associate Professor & HOD, Dept. of Psychiatry, RDJM Medical College, Muzaffarpur
• Ex-Assistant Professor, PMCH Patna
• Consultant, Healthy Minds India, Patna

Written

October 2026

Format

Research explained

Evidence

Basic science study: single-neuron recordings in 112 surgical epilepsy patients, replicated in two datasets; clinical sections are the author's labelled hypotheses

Sources

4 references

• The sections on dementia, epilepsy, anxiety and brain stimulation are the author's own interpretation and are clearly marked as hypotheses; they are not conclusions of the study's authors.
• Dr. Saurabh Kumar's research focuses on brain stimulation treatments in psychiatry, and he has completed courses in medical neuroscience (Duke University) and sleep neurobiology (University of Michigan). He is the author of The Indian Mind.

© 2026 Healthy Minds India, Patna · Research explained, October 2026