Insight | Sleep before and after learning: two windows that decide what you remember

The Short Version

Sleep plays two separate parts in learning, and the sources agree that both matter. Sleep before learning clears space in the hippocampus, the brain's short-term memory store, so it can take in new material: in Matthew Walker's lab, people who skipped a night formed about 40 percent fewer new memories, and a 90-minute nap restored the ability to learn [1]. Sleep after learning is where the brain actually saves and rewires what you studied. Sleep spindles and deep non-REM slow waves move memories from the hippocampus into the cortex, and REM sleep late in the night links new material to old knowledge, while (according to Gina Poe) it also wipes the hippocampus clean for the next day [1][2]. Andrew Huberman builds his study protocols on the same two-part idea: "you trigger the change and in sleep you get the change" [5]. For students and anyone learning a skill, the nights around the learning are part of the studying.

Table of contents

What They Said

Matthew Walker: three benefits, two of them about timing

In his conversation with Huberman, Walker, a sleep scientist at UC Berkeley and the author of Why We Sleep, sorts the science into three benefits. Sleep before learning prepares the brain to take in new information. Sleep after learning saves and strengthens it. And sleep blends new memories with everything already stored, which produces insight and creativity [1]. This article is mostly about the first two, but the third comes from the same night's work.

The night before: clearing the inbox. Walker's lab tested whether an all-nighter makes sense. Healthy volunteers either slept or stayed awake all night, then tried to learn new facts inside a brain scanner. The sleep-deprived group had a 40 percent deficit in forming new memories, and replications with different material have found deficits of 20 to 40 percent [1]. The difference showed up in the hippocampus, which Walker calls the brain's informational inbox. In rested people it lit up during learning. In sleep-deprived people there was almost no signal, as if new files were bouncing back [1]. Animal studies point the same way: sleep restriction makes the hippocampus's synapses resistant to forming new connections [1].

His lab also ran the reverse test. Two groups learned a large set of facts at midday, then learned new material six hours later, at 6 p.m. The group that stayed awake learned worse in the evening. The group that took a 90-minute nap was restored and even did slightly better, a gap of about 20 percent [1]. The benefit tracked non-REM sleep, and especially sleep spindles, short bursts of electrical activity [1]. Walker's image is a USB stick with limited space: sleep moves files to the cortex, the hard drive, so you wake with room to learn again [1].

The night after: hitting save. Walker describes sleep after learning as hitting the save button [1]. He points to a landmark study by Jenkins and Dallenbach, which he dates to 1929. It tested only two participants but has been replicated many times. Nonsense syllables were forgotten quickly across waking hours, but after two to three hours of sleep the memories stopped decaying, as if set in amber [1]. There is no need to rush to bed right after studying: the brain can hold new memories for about 16 hours until sleep consolidates them [1].

He describes two mechanisms for fact-based memory that work together. The first is transfer: deep non-REM slow waves, with sleep spindles riding on top like surfers, move memories from the vulnerable hippocampus to long-term storage in the cortex [1]. The second is replay. Bruce McNaughton and Matt Wilson found that the firing pattern of hippocampal neurons in rats running a maze was replayed during non-REM sleep 10 to 20 times faster. Wilson later found that in REM sleep the replay slowed to about half of waking speed [1].

Cramming as a false economy. Walker asks students to treat sleep the night before learning as an investment rather than lost time [1]. Staying up to study does help somewhat the next day. But in unpublished data from his lab, people tested a month later kept far more if they had slept, while the sleep-deprived group had lost almost everything [1]. Huberman calls this the cramming effect: the information never reaches long-term memory [1].

Skills get better overnight, not just safer. A pianist once told Walker that pieces he struggled with at night came easily the next morning, and that remark started about a decade of research [1]. Walker's lab had people practise a finger-tapping sequence (4-1-3-2-4) over 12 trials. After 12 hours awake, performance was unchanged. After a night of sleep, speed improved about 20 percent and accuracy almost 37 percent [1]. The awake group improved once they slept the following night, and a nap gave the same benefit, which rules out a time-of-night effect [1]. His conclusion is that it is not practice that makes perfect but "practice with a night of sleep that makes or leads to perfection" [1].

The difference between facts and skills matters. For facts, sleep prevents forgetting. For motor skills, it actually makes performance better [1]. The gains tracked stage 2 non-REM sleep spindles. In high-density EEG recordings, the spindles increased specifically over the brain region that controls the hand that had practised, which Walker likens to a masseuse working on the muscles that need it most [1]. Huberman links this to synaptic tagging, and the spindles fire about 12 to 15 times per second, a rate well suited to strengthening synapses [1]. More complex two-handed sequences gained even more, and sleep specifically smoothed the awkward pauses ("chunking") inside a sequence, building automaticity [1].

Stage 2 spindles are densest in the last quarter of the night, which is exactly the sleep that early alarms cut off. Walker suggests that coaches who wake athletes at 5 a.m. may be costing them skill gains [1]. He mentions Cheri Mah's Stanford sleep-extension study, in which basketball players improved their shooting and sprint speed [1], and he calls sleep "probably the greatest legal performance enhancing drug that most athletes are not abusing enough" [1].

School and hospital schedules. Walker takes the science into policy. He calls the typical American school start time of around 7:30 to 7:45 a.m. "lunacy, since buses leave as early as 5:30 and some children wake around 5 a.m. [1]. When Edina, a suburb of Minneapolis, moved its start from 7:25 to 8:30, the average SAT score of its top 10 percent of students reportedly rose from 1288 to 1500. Walker himself notes that some have questioned how reliable that data is [1]. Broader data consistently show better grades, fewer psychiatric problems and less truancy with later starts [1]. His summary: When sleep is abundant, minds flourish" [1]. On medicine, citing Charles Czeisler's work, he says residents on 30-hour shifts make about 460 percent more diagnostic errors in intensive care [1]. He adds that sleep loss harms decision-making as well as memory, because the frontal lobe is especially vulnerable [1].

Gina Poe: what each stage of the night does

Gina Poe, a professor of integrative biology and physiology at UCLA, studies which brain activity patterns in which sleep phases shape learning, memory, emotion and hormone release [2]. Her main point is that each stage does a different job, and that timing matters as much as how long you sleep [2].

The map of a night. Sleep cycles roughly every 90 minutes. The first REM period comes about 95 minutes after falling asleep and lasts 10 to 15 minutes, and a perfect night includes four or five cycles, about seven and a half to eight hours [2]. Stage two, once dismissed as a transition, is now known for sleep spindles and K-complexes. Stage three is deep slow-wave sleep, when big slow waves sweep through and clean the brain [2]. The first four hours of sleep look especially important for processing new memories, and dreams early in the night draw more on the day's experiences [2].

Spindles as the write window. Poe describes spindles as 10 to 15 hertz exchanges between the thalamus and the cortex. They mark a unique window in which the hippocampus, like a computer's RAM, writes to the cortex, the hard disk [2]. She cites work by Sidarta Ribeiro showing that in rats each successive REM period moved a memory further from the hippocampus through a chain of cortical areas [2]. Spindle density correlates with intelligence, and an increase in spindles on the night after learning almost perfectly predicts how well the new information is taken in [2]. People with schizophrenia have very few spindles, which may leave new information unconnected to what they already know [2]. Work by Julie Seibt and Anita Luthi shows large calcium surges and plasticity during spindles in the distal dendrites, the parts of neurons that listen to other cortical areas and the hippocampus rather than to the outside world. Poe says that is where schemas get sewn together [2].

Schemas and late-night REM. Later cycles bring more REM, and this is when schemas are built and creativity happens [2]. A schema is a web of linked ideas, like everything we associate with Christmas. Given a list of related words without the word "Christmas," people later falsely remember seeing it [2]. Huberman compares schemas to a computer desktop of meaningful folders. Poe agrees and adds that in late-night REM the brain opens the folders and compares documents, linking things that are slightly different, which is likely one origin of creativity [2].

Why REM also readies you for tomorrow. Poe's most distinctive contribution concerns the locus coeruleus, a brainstem nucleus packed with norepinephrine-containing neurons [2]. During sleep its firing slows from about two cycles per second to about one, and REM sleep is the only time it switches off completely [2]. Poe believes that silence is essential for erasing and weakening synapses that no longer serve us, especially for clearing the hippocampus's novelty-encoding pathways once memories have been moved to the cortex [2]. In her analogy, the hippocampus is a thumb drive we carry all day. During sleep we write it to the cortex's hard drive, and in REM we wipe the thumb drive so we can keep learning throughout life [2]. Norepinephrine helps strengthen synapses but blocks the weakening that lifelong learning also needs, and serotonin, which biases the brain toward novelty, also switches off in REM. Both have to be absent for the erasure to happen [2].

This is the same pattern Walker describes, seen from a different angle. For Walker, sleep before learning empties the inbox. For Poe, the REM of last night is part of what makes today's learning possible [1][2].

Consistent bedtimes. Poe also argues that when you sleep matters. The deep slow-wave sleep of the very first cycle brings a big release of growth hormone, in men and women equally, along with a push for protein synthesis that helps build memories [2]. You cannot simply move that first phase later, because every cell's circadian clock expects growth hormone at a particular time relative to melatonin [2]. Early sleep, especially the first 20 minutes, rebuilds ATP from adenosine, and slow waves then clear the day's debris, which Poe likens to cleaning up after a party so the next day's party isn't clogged [2]. Go to bed much later and you miss it, because late sleep is dominated by stage two and REM [2]. In her words, "one of the best markers of good neurological health when we get older is consistent bed times" [2]. She adds that alcohol degrades early sleep, suppressing REM and the stage-two spindles [2].

Teenagers. Like babies, teenagers are developing fast, and sleep does the hard work of organizing their brains. There is a developmental clock as well as a daily one, and missing sleep can mean missing a developmental window, so Poe's advice is to let them sleep [2].

Calm before bed. Poe advises going to sleep as calm as possible. Avoid exciting, stressful novelty such as intense video games before bed, and use deep breathing, meditation, a warm bath or a comforting book: nothing too exciting and nothing too boring [2]. Waking from deep slow-wave sleep causes sleep inertia, which she compares to opening a washing machine mid-cycle and trying to wear the soaking clothes. It is better to let the roughly 90-minute cycle finish [2]. She does not use an alarm clock unless she has a plane to catch [2].

Andrew Huberman: turning the two windows into a study protocol

Huberman does not run sleep experiments himself in these sources. He takes the sleep science and builds learning protocols on it, across three solo episodes [3][4][5].

Learning is a two-step process. In his episode on studying, Huberman says learning has two steps. The first is focus, which requires alertness, and the foundation for it is a good night's sleep before learning, along with sensible stress levels, hydration and an appropriate amount of caffeine [3]. He calls sleep the best cognitive enhancer there is [3]. The second step is the rewiring itself, which happens not while you study but during deep sleep and sleep-like states, particularly REM sleep in the later part of the night [3]. Research on the first-night effect shows that material learned on a given day is mostly consolidated during that same night's sleep [3]. A bad night does not erase what you learned, but he urges students to protect the night after learning and to be wary of late caffeine and all-nighters, which undermine the very learning they are meant to serve [3].

In his episode on focus he states the same thing more bluntly. Plasticity "doesn't occur during wakefulness. It occurs during sleep" [4]. Focused learning marks the relevant synapses, and during sleep that night and on the following nights those circuits strengthen while others are lost [4]. From the transcript: "There's a stamp in the brain where the CETO Culling was released. It actually marks those synapses norechemically and metabolically so that those synapses are more biased to change. Now if you don't ever get that deep sleep then you probably won't get those changes" [4]. (The transcription garbles "acetylcholine" there.) His conclusion: "mastering sleep is key in order to reinforce the learning that occurs" [4].

In the Essentials episode he puts it in one line: "the trigger for plasticity and learning occurs during high focus, high alertness states, not while you're asleep" [5]. Then: "So you trigger the change and in sleep you get the change" [5]. And the reverse: "if we cannot sleep well and we can't rest well, we will not access plasticity and rewire our brain because that's when the actual configuration between the connections occurs" [5].

Sleep before learning, in Huberman's terms. Huberman's version of Walker's "before window is about alertness. Alertness depends on epinephrine released from the locus coeruleus: So, no alertness, no neuroplasticity [4]. When he lists how to get alert, sleep comes first: most people rely on a cup of coffee and a good night's sleep, and he says to master your sleep schedule first, since the ability to focus tracks how well you sleep [4]. In his own words from the transcript: So get your sleep handled. But once that's in place, the question then is how do I access this alertness?" [4]. He adds that alertness is necessary but not enough, and that you can't just play material while asleep and learn it [4].

Rest right after learning. Huberman adds a step between the study session and the night: non-sleep deep rest (NSDR) or a short nap. He cites a Cell Reports paper on a hard sequence task in which people had to remember lights flashing in sequences of up to 15 or 16 [4]. In his words, "If immediately after and it was immediately after the learning the actual performance of this task people took a 20 minute non-sleep deep rest protocol or took a shallow nap so lying down feet slightly elevated perhaps just closing their eyes no sensory input. The rates of learning were significantly higher" [4]. He concludes that "you can actually accelerate learning with these NSDR protocols or with brief naps 90 minutes or less" [4]. He describes NSDR as a way to "bypass the need for deep sleep at least partially" [4]. In his studying episode he describes it as a 10 to 20 minute practice, sometimes associated with yoga nidra, that restores mental and physical vigour and enhances neuroplasticity. It can be done after a poor night, in the afternoon or in the middle of the night [3].

He also recommends easy, low-focus movement after a learning bout, such as walking, running or cycling, which creates self-generated optic flow [4]. From the transcript: "letting the mind drift where it's not organized in thought after a period of very deliberate focused effort is the best way to accelerate learning and depth of learning" [4].

The full formula. His summary for adults: "The key to plasticity in adulthood is to engage alertness engage focus and then to engage non-sleep deep rest and deep sleep while you're in your typical bout of sleep" [4]. He adds a reassurance: "let's say you get a really poor night of sleep after about of learning chances are if you sleep the next night or the following night that learning will occur" [4]. Once it is consolidated, he says, "you will have acquired the knowledge forever unless you go through some process to actively unlearn it" [4].

Daily timing. In the Essentials episode, Huberman lays out a day built around these windows. The first three hours after waking are the most alert for most people [5]. He gets morning light, delays caffeine about two hours, exercises early, then does a focused learning bout [5]. In the afternoon, when he gets groggy, he uses NSDR rather than a double espresso: "I personally find it a mistake to that point down a double espresso and charge really hard. It just doesn't work for me. I end up really disrupting my sleep schedule" [5]. His second, afternoon bout is done without caffeine "because I want to make sure that I can sleep later that night really well" [5]. He dims lights and minimizes bright overhead light from about 10 p.m. to 4 a.m. to protect the sleep where the rewiring happens [5].

Where They Agree, and Where They Don't

Clear agreement

• Sleep is active work on memory, not downtime. Walker's closing message is that sleep is active information processing [1]. Poe says every stage does a different, essential job [2]. Huberman places the actual rewiring in sleep [3][4][5].
• The hippocampus is a limited short-term store that sleep empties into the cortex. All three use the same computer comparison. Walker: a USB stick moved to a hard drive [1]. Poe: RAM writing to the hard disk, and a thumb drive wiped in REM [2]. Huberman builds on the same hippocampus-to-cortex picture when he says sleep strengthens the circuits you tagged during learning [4].
• Spindles and deep non-REM sleep do the transfer. Walker ties restored learning after a nap, and overnight skill gains, to spindles [1]. Poe calls spindles a unique write window and says an increase in spindles after learning almost perfectly predicts how well new information is taken in [2].
• Cramming and all-nighters work against their own goal. Walker's month-later data and Huberman's warning about all-nighters and late caffeine say the same thing [1][3].
• Naps count. Walker's 90-minute nap restored learning capacity and produced the same skill gain as a night [1]. Huberman cites naps of "90 minutes or less" and 20-minute NSDR sessions as speeding up learning [4].

Where they differ in emphasis

Which sleep stage does the saving. Walker puts most of the weight for facts on deep non-REM slow waves and spindles, and for skills on stage 2 spindles late in the night [1]. Poe gives REM a bigger role: Ribeiro's rats moved a memory further from the hippocampus with each REM period, and REM's chemical silence is what wipes the hippocampus for tomorrow [2]. Huberman's own wording shifts between episodes. In the studying episode the rewiring happens particularly in REM in the later part of the night [3]. In the focus and Essentials episodes he stresses "deep sleep" and NSDR [4][5]. These aren't really contradictions. The sources describe a sequence: spindles and slow waves move the memory, and REM links and cleans up. In practice, that means you need the whole night.

Early night or late night. Poe says the first four hours look especially important for new memories, and that a much-delayed bedtime misses the first cycle's growth hormone and brain-cleaning [2]. Walker warns that the last quarter of the night holds the densest stage 2 spindles for motor skills, and that early alarms cut it off [1]. Both are true in their own terms: going to bed late costs the early part of the night, and waking early costs the late part. So the advice is a consistent bedtime and a full night, not trading one end for the other.

How forgiving is one bad night? Here the tone differs. Huberman reassures: a bad night does not erase what you learned, and the next night or the one after can still do the work [3][4]. Walker's unpublished data are harsher: a month later, the sleep-deprived group had lost almost everything [1]. Walker's 16-hour holding window [1] also suggests the brain does not wait indefinitely. The two positions can be reconciled: Walker's data describe a full night without sleep after learning, while Huberman is talking about a poor night, not none. Walker's data are unpublished, and Huberman's reassurance does not come with a study in these sources, so neither side should be pushed too far.

Caffeine. Walker says there is some evidence that caffeine enhances hippocampal encoding, though no one has yet tested whether it rescues learning after sleep deprivation [1]. Huberman accepts an appropriate amount of caffeine as part of the foundation for focus [3], but warns about late caffeine [3] and avoids it in his afternoon bout to protect sleep [5]. Nobody here recommends caffeine as a substitute for sleep.

Sleep trackers. This is a side point, but it matters for anxious students. Walker warns about "orthosomnia," anxiety about achieving perfect sleep, often fuelled by trackers [1]. Poe wears one but doesn't live by it, since "the best ones right now are about 70% effective at staging your sleep" [2].

Why It Works

Before learning: capacity. The sources give a consistent mechanism for why a rested brain learns better. The hippocampus holds new information for a while, and its space is limited [1][2]. Sleep, through non-REM spindles, moves yesterday's material into the cortex [1][2]. REM then silences the locus coeruleus, and without norepinephrine and serotonin the brain can weaken and clear the hippocampal novelty-encoding pathways it no longer needs [2]. Walker's scanner study shows what happens when that doesn't happen: the hippocampus barely responds to new material [1]. Animal work shows sleep restriction making hippocampal synapses resistant to forming new connections [1].

There is also an alertness side. Huberman explains that learning needs epinephrine from the locus coeruleus for alertness, and acetylcholine for focus [4]. Sleep loss undermines that alertness, and Walker adds that the frontal lobe is especially vulnerable to sleep loss, which harms decision-making [1]. Poe adds a cleaning role: early slow waves rebuild ATP and clear the debris of waking activity, like a bilge pump pushing fluid through the brain [2].

After learning: transfer, replay and strengthening. Walker gives two mechanisms. Transfer means slow waves carrying spindles move memories from hippocampus to cortex [1]. Replay means hippocampal neurons re-fire the day's patterns during non-REM sleep, 10 to 20 times faster than waking [1]. Poe adds that spindles are a thalamus-cortex exchange during which the hippocampus writes to the cortex, with calcium surges and plasticity in the dendrites that listen to the hippocampus [2]. Spindles fire about 12 to 15 times per second, which is suited to strengthening synapses, and in Walker's skill study they concentrated over the brain area for the hand that had practised [1]. Huberman describes the front end of the same process: focused learning, with acetylcholine released, tags the synapses so they are "more biased to change," and sleep then strengthens them [4].

The evidence as the sources give it:

• Walker's all-nighter scanner study: a 40 percent deficit in new memory formation, replicated at 20 to 40 percent [1].
• Walker's midday nap study: about 20 percent better evening learning after a 90-minute nap than after staying awake, tracking non-REM spindles [1].
• Jenkins and Dallenbach (dated by Walker to 1929): memories stopped decaying after two to three hours of sleep. It had only two participants but has been replicated many times [1].
• McNaughton and Wilson: fast replay of maze patterns in rats during non-REM sleep [1].
• Walker's finger-tapping study: about 20 percent faster and almost 37 percent more accurate after sleep, with no change after 12 hours awake [1].
• Ribeiro's rat work: each REM period moved a memory further from the hippocampus [2].
• Spindle increases after learning almost perfectly predicting how well it is taken in [2].
• The Cell Reports paper: 20 minutes of NSDR or a shallow nap right after a hard sequence task raised learning significantly [4].
• The first-night effect: most consolidation happens that same night [3].

The weaker points should be named. Walker's month-later cramming result is unpublished [1], and he flags that the Edina SAT figures have been questioned [1]. Several mechanisms, including REM erasure and PTSD links, rest partly on animal work and Poe's hypotheses [2].

Where It Applies, and Where It Doesn't

Facts and skills, differently. For fact-based (declarative) memory, sleep after learning mainly prevents forgetting [1]. For procedural skills such as music, sport or typing, a night's sleep actually improves performance, and more complex sequences gain more [1]. So the sources apply to both classroom study and physical practice, but the payoff looks different.

Naps work, within limits. A 90-minute afternoon nap restored learning capacity and produced overnight-level skill gains in Walker's studies [1]. Huberman's source supports short naps and 20-minute NSDR right after learning [4]. None of the sources suggests naps can replace a full night, and Poe's point about the first cycle's growth hormone and cleaning depends on a regular night-time bedtime [2].

Not a shortcut to learning without effort. Sleep only consolidates what was tagged while you were awake. Huberman says alertness is necessary but not sufficient, and you can't learn material by playing it while asleep [4]. His blunter version: "One of the biggest lies in the universe that seems quite prominent right now is that every experience you have changes your brain" [4]. Sleep strengthens what you worked on with focus. It does not rescue a distracted session.

Night owls and shifted schedules. Poe says someone whose schedule is consistently late, say 1 or 2 a.m. to 10 a.m., may be fine if everything has shifted, though that should be studied [2]. The problem is inconsistency, going to bed much later than usual, not lateness itself [2]. Walker's advice to time learning to your circadian peak (about 10 to 11 a.m. for an early riser, closer to midday or 1 p.m. for someone who sleeps 11 p.m. to 7:30 a.m.) likewise adjusts to the person [1].

Teenagers and school. The sources treat adolescents as a special case. Poe says teenagers are still developing and missing sleep can mean missing a developmental window [2]. Walker argues that early school start times work directly against this, with later starts linked to better grades, fewer psychiatric problems and less truancy [1].

Adults versus children. Huberman notes that children learn almost passively, while after about 25 adults have to earn plasticity through alertness, focus and then rest and sleep: "The key to plasticity in childhood is to be a child" [4]. The two-window pattern matters most for adults deliberately trying to learn.

Health conditions. If you consistently need nine or more hours to function, Poe says it may signal inefficient sleep from problems like sleep apnea, unnoticed snoring or frequent waking, and a sleep study may be worthwhile [2]. Persistent sleep problems, or questions about medications that affect sleep, are worth taking to a doctor or sleep specialist rather than solving with study tricks.

What To Do With It

1. Sleep the night before anything important you need to learn. Treat it as an investment, not lost study time. Walker's scanner study found a 40 percent drop in new memory formation after an all-nighter [1]. Huberman calls a good night's sleep before learning the foundation of focus [3].

2. Protect the night after, and don't cram. Most consolidation happens that very night [3]. People who slept kept far more a month later than those who crammed [1]. Avoid all-nighters and late caffeine in the days you're learning [3].

3. Keep your bedtime consistent. The first sleep cycle's growth hormone, protein synthesis and brain cleaning depend on going to bed at your usual time, and a bedtime pushed two hours late misses it [2]. Skip alcohol in the hours before bed, since it suppresses REM and stage-two spindles [2].

4. Don't cut off the end of the night. Stage 2 spindles that improve motor skills are densest in the last quarter of the night [1]. If you're practising an instrument or a sport, avoid early alarms where you can, and try to wake at the end of a cycle rather than in the middle of deep sleep [1][2].

5. Study in focused bouts at your alert time. Time learning to your circadian peak [1]. Huberman suggests bouts of about 90 minutes with 5 to 10 minutes of warm-up [4], mostly in the most alert hours after waking [5]. The effort and focus are what mark the synapses for sleep to strengthen [4].

6. Rest right after a hard session. Take a 20-minute NSDR session or a shallow nap soon after demanding learning: lie down, feet slightly raised, eyes closed, no sensory input [4]. A walk or easy run without a podcast also lets the mind drift [4].

7. Use a nap to reset in the middle of a long study day. In Walker's study, a 90-minute nap between a midday and an evening session restored learning capacity [1].

8. Wind down calmly before bed. Poe recommends avoiding exciting, stressful novelty such as intense video games, and using deep breathing, meditation, a warm bath or a comforting book: nothing too exciting and nothing too boring [2]. Dim bright overhead lights late in the evening [5].

9. Don't panic after one bad night. Huberman says learning can still consolidate on the next night or the one after [4], and Walker notes the brain can hold new memories for about 16 hours until sleep [1]. Don't obsess over tracker scores either: Walker suggests checking data only about once a week [1].

10. Make it policy if you're a parent, coach or teacher. Let teenagers sleep [2], support later school start times [1], and don't schedule athletes' practices so early that they lose the last hours of sleep [1].

Sources

1. Dr Matt Walker Using Sleep To Improve Learning Creativity Memory · Huberman Lab
2. Dr. Gina Poe: Use Sleep to Enhance Learning, Memory & Emotional State · Huberman Lab
3. Optimal Protocols for Studying & Learning · Huberman Lab
4. How to Focus to Change Your Brain · Huberman Lab
5. Optimize Your Learning & Creativity With Science-Based Tools | Huberman Lab Essentials · Huberman Lab