Daily

Practical guide

How to Improve Working Memory: What Works and What Doesn't

By LaserMind Team ·

Two hands writing notes across several open notebooks
Photo: cottonbro studio on Pexels

What "improving working memory" can mean

What working memory is, and how researchers measure it, is covered in What is working memory? This guide is about the practical question. "Improving working memory" can mean four different things, and the evidence differs for each:

  • The task you practise: getting better at n-back or digit span itself.
  • Near transfer: getting better at similar working-memory tasks you have not practised.
  • Far transfer: getting better at other things that rely on working memory, such as reasoning, reading, arithmetic or keeping track of a busy day.
  • Working around the limit: using the capacity you have more efficiently, and avoiding what temporarily impairs it.

Several numbers below are effect sizes (g or d), which express the difference between two groups in a standard unit so that studies using different tests can be combined. A minus sign means the group under strain, such as sleep loss or stress, did worse.

Does working-memory training work?

Working-memory training means practising demanding memory tasks, usually on a computer and usually adaptive: the task gets harder as you improve. Common examples are n-back, span tasks and commercial programs such as Cogmed, a computerised, adaptive working-memory training program [1].

The tasks you practise improve

A major review of the studies cited by brain-training companies found extensive evidence that training improves performance on the trained tasks, less evidence for closely related tasks, and little evidence for distantly related tasks or everyday cognitive performance [2]. In a 2016 meta-analysis by Monica Melby-Lervåg, Thomas Redick and Charles Hulme, the largest effects were on tasks similar or identical to the trained ones (g = 0.80 against control groups that also did computer training) [3].

Similar tasks improve a little, for a while

That meta-analysis pooled 87 publications with 145 comparisons, covering children, adults and older adults, some selected for learning difficulties [3]. The fairest comparisons are with "treated" control groups, which also did computer training: non-adaptive memory tasks, or tasks that did not train memory. Against them, trained groups improved more on untrained verbal working-memory tasks (g = 0.31) and visuospatial ones (g = 0.28) straight after training [3]. About five months later on average, the verbal gain was no longer significant. Visuospatial gains were still there, but many of those tests shared stimuli or methods with the trained tasks [3]. An earlier meta-analysis of 23 studies found the same pattern [4].

A 2017 meta-analysis looked only at n-back training in healthy adults: 33 randomised controlled trials with 2,105 participants [5]. Transfer to untrained versions of the n-back task was medium-sized (g = 0.62); transfer to other working-memory tasks was small (g = 0.24). The authors concluded that earlier meta-analyses had overestimated transfer to working memory by counting untrained versions of the training task as working-memory tests [5].

Other abilities: no convincing improvement

Against treated controls, the 2016 meta-analysis found no convincing evidence of improvement in nonverbal ability, verbal ability, word decoding, reading comprehension or arithmetic, and the size of a study's working-memory gain was not related to the size of its far-transfer effect [3]. Two small effects, on reading comprehension straight after training and on arithmetic at follow-up, came mainly from studies in which the control group's scores fell, and the authors judged that they were not evidence of a training effect [3]. In the n-back meta-analysis, effects on fluid intelligence (reasoning) tests and cognitive-control tasks were very small (g = 0.16 each), and neither remained significant once the type of control group was taken into account [5].

The research is not unanimous. A 2015 meta-analysis of 20 n-back studies in healthy adults aged 18 to 50 found a small but significant improvement on fluid intelligence tests [6]. A reanalysis that adjusted for baseline differences produced a smaller effect (g = 0.13, against the original 0.24), which lost significance when only studies with active control groups were kept; the original authors replied that there was no evidence that the type of control group itself changed the effect [7]. Looking across meta-analyses of many kinds of cognitive training, Sala and Gobet found minimal effects on general cognitive skills, with differences between studies explained by design quality and statistical artefacts [8].

Cogmed-style programs

A meta-analysis of 50 studies of Cogmed found highly consistent near-zero effects on far-transfer measures such as attention, intelligence, language and mathematics. Effects on memory tasks were small to medium, and weaker for tasks not directly related to the trained ones [9].

The largest single trial screened 1,723 children aged 6 to 7 in 44 schools in Melbourne, Australia, and randomly assigned 452 children with low working memory either to Cogmed (20 to 25 sessions of 45 minutes at school) or to usual classroom teaching [1]. Visuospatial short-term memory, one of four memory outcomes, improved at 6 and 12 months (effect sizes 0.43 and 0.49) but not at 24 months. No other outcome improved, and the trained children's maths scores were worse at two years [1].

Why the gains stay close to the task

One explanation is strategy: people learn tricks that suit the task they practise. The authors of the n-back meta-analysis suggested that its transfer comes mainly from task-specific aspects, such as suitable strategies [5].

A pre-registered trial with 116 adults aged 18 to 40 taught one group a strategy for a digit n-back task: arrange the incoming digits in your mind into two rows, N digits long, one above the other, and compare each pair [10]. After a single 30-minute session, this group reached higher n-back levels than a group that trained without the strategy and a group that did no training. Compared with the uninstructed group, it also did better on harder untrained n-back tasks, but not on structurally different working-memory tasks such as forward digit span [10].

OutcomeWhat the meta-analyses found
The task you practiseReliable, large gains
Untrained versions of the same taskMedium gains after n-back training
Other working-memory tasksSmall gains straight after training; verbal gains faded by follow-up
Reasoning, reading, arithmeticNo convincing improvement against control groups that also trained

If you enjoy n-back, the n-back method page explains how to practise it, and you can play N-Back online. Expect to get better at n-back.

Use the capacity you have more efficiently

Chunk information into familiar units

Chunking means recoding small units into larger, familiar ones, such as splitting a phone number into groups. In four laboratory experiments, familiar chunks were recalled better and also improved recall of other information held at the same time [11]. The authors concluded that a chunk replaces its separate elements with one compact representation retrieved from long-term memory, which frees capacity for what comes next. Chunks helped other material only when they came early in a list, and a chunk has to be something you already know [11].

Practice and strategy can push a span score a long way. In a classic study of one volunteer, more than 230 hours of laboratory practice raised his digit span from 7 to 79 digits, and the authors credited an appropriate mnemonic system [12]. LaserMind grades chunking E3 (limited evidence) for encoding long sequences.

Write it down when you only need to use it

Researchers call it cognitive offloading when you use a physical action, such as writing a note or setting a phone reminder, to reduce the mental demands of a task [13]. In a registered study, 114 undergraduates heard lists of 2 to 10 letters and repeated them in order [14]. When they could write the letters down, they did better than when they relied on memory alone, especially with longer lists. Students with lower working-memory capacity were not more likely to write things down, and did not benefit more [14].

Offloading has a cost. In three experiments with 172 students each, participants copied patterns of images from a model to a workspace [15]. The more they relied on looking back at the model, the faster they worked, but the less they remembered of the patterns in a later test. In the second experiment, offloading still harmed memory when people knew a test was coming. In the third, people who were made to offload as much as possible, but knew about the test, almost completely avoided the cost [15].

So offload what you only need to use, such as a list of steps, a phone number or tomorrow's tasks. When the goal is to learn the material, decide to remember it rather than relying on the note.

Cut out load that doesn't help

Cognitive load theory separates the load that comes from the material itself from extraneous load, which comes from how information is presented and what you are asked to do, and which can be changed by changing the instruction [16]. For example, a meta-analysis of 58 comparisons with 2,426 participants found that placing related words and pictures close together improved learning (g = 0.63) [17]. In everyday terms: keep what belongs together in one place, remove what you don't need, and take one step at a time. What is cognitive load? explains the theory and its limits.

Protect it: sleep, stress and exercise

The sleep and stress experiments below measured what impairs working memory for a while. They do not show that more sleep or less stress than usual raises it above your normal level.

Sleep

A 2010 meta-analysis of 70 articles looked at healthy adults kept awake for up to 48 hours (1,533 participants in total) [18]. Accuracy on working-memory tasks fell (g = −0.555), and so did speed (g = −0.515); the largest effects were on simple attention, and reasoning accuracy barely changed [18]. Shorter nights matter too: in a meta-analysis of 61 studies in which people were allowed less sleep rather than none, cognition was impaired overall (g = −0.38), including executive functioning (g = −0.32) [19]. See Sleep and memory for more.

Acute stress

A 2016 meta-analysis pooled 34 studies with 1,353 participants without a known psychiatric disorder, who did working-memory tasks after a brief laboratory stressor, such as giving a speech in front of judges [20]. Stress impaired working memory with a small average effect (g = −0.197). The effect was larger when the task placed a high load on working memory (g = −0.303) than when it did not (g = −0.049, not significant) [20]. How does stress affect attention and memory? covers what helps in the moment and how strong that evidence is.

Exercise

An umbrella review of 133 systematic reviews, covering all ages and including clinical groups, found that exercise improved executive function, the category in which it counted working-memory tests (standardised mean difference 0.24); 94 of the 133 reviews were rated critically low in quality [21]. A stricter umbrella review, limited to randomised trials in healthy people, found a small benefit for cognition (d = 0.22) that shrank once active control groups and baseline differences were accounted for (d = 0.13), and became negligible after correcting for publication bias (d = 0.05) [22].

That conflict is why LaserMind grades aerobic exercise E2 (conditional evidence) for general cognition, memory and executive function. Exercise for your health first, and treat any benefit for working memory as a possible bonus; How does exercise affect brain health and cognition? goes through the evidence.

What doesn't work, or isn't established

  • Expecting a memory game to improve everything else. Training has not convincingly improved reasoning, reading, maths or general cognitive skills [3][8].
  • Reading a higher score as a bigger memory. Part of the rise on a practised task comes from strategies and familiarity with that task [5][10].
  • Supplements. This article does not cover them.

How to improve working memory, step by step

  1. Chunk. Group numbers, letters and items into units you already know.
  2. Offload what you only need to use. Write down lists, numbers and next steps. When you need to learn something, decide to remember it.
  3. Strip out extraneous load. Keep related information together and work through one step at a time.
  4. Protect the conditions. When you can, avoid demanding mental work after a night of little or no sleep, and settle down before high-load tasks.
  5. Keep moving for your health. Treat any effect on thinking as a bonus.
  6. Practise working-memory games for their own sake, and track them honestly, as the next section explains.

How to practise on LaserMind honestly

LaserMind has practice versions of three classic tasks:

  • N-Back: a square lights up on a 3×3 grid every 3 seconds, and you press Match when it is in the same place as N steps back. After each of the three blocks, N goes up if you got at least 90% right and down if you got 70% or less. Your main number is the highest N you played with at least 80% correct. A Dual mode adds a spoken letter.
  • Digit Span: digits appear one at a time, about a second each, and you type them back with no time limit. The sequence starts at 3 digits; you get two tries at each length, and it grows by one digit when at least one try is right. Backward mode asks for the reverse order, and Chunked mode shows the same digits in groups of three.
  • Light Path: nine blocks in a fixed, irregular layout light up one at a time, and you tap the same path, or its reverse in Backward mode. The path starts at 2 blocks and can reach 9, with the same two-tries rule.

Forward span is closer to short-term memory than to working memory; Working memory vs short-term memory explains the difference. LaserMind grades n-back E2 (conditional evidence) for n-back and similar working-memory tasks, digit span E2 for trained span tasks, and Corsi block tapping, the task behind Light Path, E2 for spatial span tasks. The grades describe what practice improves: the tasks themselves.

To practise honestly:

  • Expect your scores to rise, and know why. A meta-analysis of 50 studies (134,436 people) found an average gain of 0.26 on cognitive ability tests that were simply taken again, larger when the identical test was reused [23]. Strategies add more.
  • Watch a strategy at work. Play Digit Span Forward, then Chunked, and compare the two spans. The difference shows what grouping does, not a change in capacity.
  • Compare yourself with yourself. Progress reports compare your results with your own history, for each exercise and version. Optional leaderboards for adults rank practice scores on one exercise, mode and device type; they are not a measure of everyday memory.
  • Check transfer with tasks you don't train. About every four weeks, the Check-in runs four short tasks in versions that LaserMind's programs and Daily Circuit don't train, including Digit Span and Light Path backwards. If you practise those backward modes in your own training, your report notes that they are no longer untrained.
  • Judge daily life by daily life. The test is whether keeping track of things at work or in study gets easier, not whether your n-back level rises.

Common mistakes

  • Treating a game score as a measure of your memory. It measures your practice on that game.
  • Offloading what you need to learn. Relying on notes can leave less in memory [15].
  • Training while exhausted or under acute stress. Both impair performance on working-memory tasks [18][20].
  • Judging a program by its first weeks. Early gains on a repeated task are partly familiarity [23].

What the research says

◆ SupportedPractising working-memory tasks improves performance on them, and straight after training on similar untrained working-memory tasks [3][2][5]. Up to 48 hours without sleep impairs accuracy and speed on working-memory tasks [18], and acute laboratory stress impairs working memory, with a small average effect [20].
◇ ConditionalWhether near-transfer gains last: verbal gains were no longer significant about five months after training, and the visuospatial gains that persisted were measured with tasks similar to the trained ones [3]. Chunking freed capacity in laboratory experiments when the chunks were familiar and came early [11]. Writing items down improved performance on a memory task [14], but relying on offloading reduced later memory for the offloaded material, and knowing about a later test removed that cost in only one of two experiments that tested it [15]. Exercise improved executive-function scores on average across many populations [21], but the benefit in healthy people was negligible after correcting for publication bias [22].
✕ Not establishedThat working-memory training improves reasoning, reading, arithmetic or general cognitive ability [3][8][9]; a small effect on fluid intelligence tests after n-back training is disputed [6][7]. That practising LaserMind's exercises improves working memory in daily life.

Try it on LaserMind

▶ Number Span

Play Forward, then Chunked, and compare the two spans with your own earlier results.

Try it on LaserMind

▶ N-Back

Practise n-back for its own sake: your level shows your progress on this task.

What research has not established

  • Most training studies lasted weeks and used laboratory tests. The largest meta-analysis mixed children, adults and groups with learning difficulties [3], the n-back meta-analysis covered healthy adults only [5], and the large Cogmed trial covered 6- and 7-year-olds with low working memory [1].
  • Meta-analyses disagree about fluid intelligence tests, partly because of how they handle baseline differences and control groups [7].
  • The strategy trial was one study with one 30-minute session, and its strategy group expected to improve more than the passive control group [10]. The 79-digit span came from one person [12]. The chunking and offloading experiments used short laboratory tasks, mostly with students [11][14][15].
  • The sleep and stress studies used strong manipulations, such as a night or more without sleep or a speech in front of judges [18][20]. The exercise evidence conflicts: one umbrella review included clinical groups and mostly low-quality reviews [21], while a stricter one in healthy people found the benefit negligible after correcting for bias [22].
  • LaserMind's exercises run in a browser, are not normed or diagnostic, and have not been tested in trials of their own. Tracking your own results has no control group, so it cannot separate practice from real change.

Frequently asked questions

Can you increase your working-memory capacity? Practice raises your scores on the tasks you practise, and similar tasks improve a little straight after training [3]. Whether capacity itself grows is not established, and part of the gain comes from task-specific strategies [5]. Chunking and offloading let you do more with the capacity you have [11][14].

Is dual n-back better than single n-back? In the n-back meta-analysis, training type (single or dual) did not change the results [5]. Choose the version you find engaging, and expect to get better at that version.

I keep losing track of things. Should I worry? Everyone's working memory is small, and sleep loss and stress make it worse for a while [18][20]. If forgetfulness is new, getting worse, or affecting your work or daily life, talk to a doctor or another qualified professional; online exercises cannot assess it.

What the research says

Last evidence check 2026-10-19
ⓘ Background

The n-back task was introduced in the 1950s to study short-term memory for rapidly changing information. Trained task [24]

Conditional evidence

A meta-analysis of working-memory training found reliable short-term gains on working-memory tasks, which did not last well, and no convincing transfer to verbal or nonverbal ability, word decoding or arithmetic. Trained & similar tasks [4]

Not supported

Across meta-analyses, cognitive training did not improve general cognitive ability once study quality and placebo effects were taken into account. Other abilities [8][25]

ⓘ Background

A classic paper on the limits of short-term memory argued that recoding input into chunks lets people stretch a tight limit on how much they can hold at once. Trained task [26]

Limited evidence

In four lab experiments, sequences that formed familiar chunks were recalled better and also left more room for other items held at the same time. Trained task [11]

Conditional evidence

An umbrella review of 133 systematic reviews found that exercise, including aerobic and resistance training, improved general cognition, memory and executive function; effects were small, varied between groups, and most of the reviews were rated critically low in quality. Other abilities [21]

Not supported

In healthy people, a stricter umbrella review of randomised trials found the benefit negligible after correcting for publication bias, and a Cochrane review and a two-year trial of 1,635 older adults found no cognitive benefit. Other abilities [22][27][28]

Conditional evidence

In a one-year trial with older adults, walking three times a week increased the size of the hippocampus by about 2%. Spatial memory improved about as much in the stretching control group; within the walking group, a larger increase went with a larger memory gain. Health [29]

References

  1. Roberts, G., Quach, J., Spencer-Smith, M., Anderson, P. J., Gathercole, S., Gold, L., Sia, K.-L., Mensah, F., Rickards, F., Ainley, J., & Wake, M. (2016). Academic Outcomes 2 Years After Working Memory Training for Children With Low Working Memory: A Randomized Clinical Trial. JAMA Pediatrics, 170(5), e154568. https://doi.org/10.1001/jamapediatrics.2015.4568
  2. Simons, D. J., Boot, W. R., Charness, N., Gathercole, S. E., Chabris, C. F., Hambrick, D. Z., & Stine-Morrow, E. A. L. (2016). Do "brain-training" programs work? Psychological Science in the Public Interest, 17(3), 103–186. https://doi.org/10.1177/1529100616661983
  3. Melby-Lervåg, M., Redick, T. S., & Hulme, C. (2016). Working Memory Training Does Not Improve Performance on Measures of Intelligence or Other Measures of “Far Transfer”: Evidence From a Meta-Analytic Review. Perspectives on Psychological Science, 11(4), 512–534. https://doi.org/10.1177/1745691616635612
  4. Melby-Lervåg, M., & Hulme, C. (2013). Is working memory training effective? A meta-analytic review. Developmental Psychology, 49(2), 270–291. https://doi.org/10.1037/a0028228
  5. Soveri, A., Antfolk, J., Karlsson, L., Salo, B., & Laine, M. (2017). Working memory training revisited: A multi-level meta-analysis of n-back training studies. Psychonomic Bulletin & Review, 24(4), 1077–1096. https://doi.org/10.3758/s13423-016-1217-0
  6. Au, J., Sheehan, E., Tsai, N., Duncan, G. J., Buschkuehl, M., & Jaeggi, S. M. (2015). Improving fluid intelligence with training on working memory: a meta-analysis. Psychonomic Bulletin & Review, 22(2), 366–377. https://doi.org/10.3758/s13423-014-0699-x
  7. Au, J., Buschkuehl, M., Duncan, G. J., & Jaeggi, S. M. (2016). There is no convincing evidence that working memory training is NOT effective: A reply to Melby-Lervåg and Hulme (2015). Psychonomic Bulletin & Review, 23(1), 331–337. https://doi.org/10.3758/s13423-015-0967-4
  8. Sala, G., & Gobet, F. (2019). Cognitive training does not enhance general cognition. Trends in Cognitive Sciences, 23(1), 9–20. https://doi.org/10.1016/j.tics.2018.10.004
  9. Aksayli, N. D., Sala, G., & Gobet, F. (2019). The cognitive and academic benefits of Cogmed: A meta-analysis. Educational Research Review, 27, 229–243. https://doi.org/10.1016/j.edurev.2019.04.003
  10. Laine, M., Fellman, D., Waris, O., & Nyman, T. J. (2018). The early effects of external and internal strategies on working memory updating training. Scientific Reports, 8(1), 4045. https://doi.org/10.1038/s41598-018-22396-5
  11. Thalmann, M., Souza, A. S., & Oberauer, K. (2019). How does chunking help working memory? Journal of Experimental Psychology: Learning, Memory, and Cognition, 45(1), 37–55. https://doi.org/10.1037/xlm0000578
  12. Ericsson, K. A., Chase, W. G., & Faloon, S. (1980). Acquisition of a Memory Skill. Science, 208(4448), 1181–1182. https://doi.org/10.1126/science.7375930
  13. Risko, E. F., & Gilbert, S. J. (2016). Cognitive Offloading. Trends in Cognitive Sciences, 20(9), 676–688. https://doi.org/10.1016/j.tics.2016.07.002
  14. Morrison, A. B., & Richmond, L. L. (2020). Offloading items from memory: individual differences in cognitive offloading in a short-term memory task. Cognitive Research: Principles and Implications, 5(1), 1. https://doi.org/10.1186/s41235-019-0201-4
  15. Grinschgl, S., Papenmeier, F., & Meyerhoff, H. S. (2021). Consequences of cognitive offloading: Boosting performance but diminishing memory. Quarterly Journal of Experimental Psychology, 74(9), 1477–1496. https://doi.org/10.1177/17470218211008060
  16. Sweller, J., van Merriënboer, J. J. G., & Paas, F. (2019). Cognitive Architecture and Instructional Design: 20 Years Later. Educational Psychology Review, 31(2), 261–292. https://doi.org/10.1007/s10648-019-09465-5
  17. Schroeder, N. L., & Cenkci, A. T. (2018). Spatial Contiguity and Spatial Split-Attention Effects in Multimedia Learning Environments: a Meta-Analysis. Educational Psychology Review, 30(3), 679–701. https://doi.org/10.1007/s10648-018-9435-9
  18. Lim, J., & Dinges, D. F. (2010). A meta-analysis of the impact of short-term sleep deprivation on cognitive variables. Psychological Bulletin, 136(3), 375–389. https://doi.org/10.1037/a0018883
  19. Lowe, C. J., Safati, A., & Hall, P. A. (2017). The neurocognitive consequences of sleep restriction: A meta-analytic review. Neuroscience & Biobehavioral Reviews, 80, 586–604. https://doi.org/10.1016/j.neubiorev.2017.07.010
  20. Shields, G. S., Sazma, M. A., & Yonelinas, A. P. (2016). The effects of acute stress on core executive functions: A meta-analysis and comparison with cortisol. Neuroscience & Biobehavioral Reviews, 68, 651–668. https://doi.org/10.1016/j.neubiorev.2016.06.038
  21. Singh, B., et al. (2025). Effectiveness of exercise for improving cognition, memory and executive function: A systematic umbrella review and meta-meta-analysis. British Journal of Sports Medicine. https://doi.org/10.1136/bjsports-2024-108589
  22. Ciria, L. F., Román-Caballero, R., Vadillo, M. A., Holgado, D., Luque-Casado, A., Perakakis, P., & Sanabria, D. (2023). An umbrella review of randomized control trials on the effects of physical exercise on cognition. Nature Human Behaviour, 7(6), 928–941. https://doi.org/10.1038/s41562-023-01554-4
  23. Hausknecht, J. P., Halpert, J. A., Di Paolo, N. T., & Moriarty Gerrard, M. O. (2007). Retesting in selection: A meta-analysis of coaching and practice effects for tests of cognitive ability. Journal of Applied Psychology, 92(2), 373–385. https://doi.org/10.1037/0021-9010.92.2.373
  24. Kirchner, W. K. (1958). Age differences in short-term retention of rapidly changing information. Journal of Experimental Psychology, 55(4), 352–358. https://doi.org/10.1037/h0043688
  25. Sala, G., Aksayli, N. D., Tatlidil, K. S., Tatsumi, T., Gondo, Y., & Gobet, F. (2019). Near and Far Transfer in Cognitive Training: A Second-Order Meta-Analysis. Collabra: Psychology, 5(1), 18. https://doi.org/10.1525/collabra.203
  26. Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63(2), 81–97. https://doi.org/10.1037/h0043158
  27. Young, J., Angevaren, M., Rusted, J., & Tabet, N. (2015). Aerobic exercise to improve cognitive function in older people without known cognitive impairment. Cochrane Database of Systematic Reviews, 2015(4). https://doi.org/10.1002/14651858.CD005381.pub4
  28. Sink, K. M., Espeland, M. A., Castro, C. M., Church, T., Cohen, R., Dodson, J. A., Guralnik, J., Hendrie, H. C., Jennings, J., Katula, J., Lopez, O. L., McDermott, M. M., Pahor, M., Reid, K. F., Rushing, J., Verghese, J., Rapp, S., & Williamson, J. D. (2015). Effect of a 24-Month Physical Activity Intervention vs Health Education on Cognitive Outcomes in Sedentary Older Adults: The LIFE Randomized Trial. Jama, 314(8), 781. https://doi.org/10.1001/jama.2015.9617
  29. Erickson, K. I., Voss, M. W., Prakash, R. S., et al. (2011). Exercise training increases size of hippocampus and improves memory. PNAS, 108(7), 3017–3022. https://doi.org/10.1073/pnas.1015950108