How to use Sleep Sage

Sleep Sage turns your slides or notes into short spoken cues: key terms from your material, each with a one-line explanation. You go through the cues while you are awake, and your phone can play them very quietly at night on a timer you set.

Sleep Sage is a study tool. It is not a medical device or a health service: it does not diagnose, treat, cure, or prevent any disease or sleep disorder, and it does not measure or track your sleep. It has not been tested in a scientific study, and it makes no claim about memory, learning, sleep, or test results.

The steps below borrow their set-up from published studies of sounds played during sleep, known as targeted memory reactivation. Most of those studies were small, most were run in sleep laboratories, and their results vary. The notes describe what the researchers did, not what Sleep Sage will do for you. If you have trouble sleeping or concerns about your memory, talk to a health professional.

1. Study your material first

   Upload a lesson, then study it the way you normally would. Sleep Sage picks key terms printed in your material, so each cue is a word you have already met while studying.

   What the studies did: In these studies, every sound played at night had first been paired with something people learned while awake, including in one study where students studied at home. The sounds were never new to them. (Rudoy et al., 2009; Schreiner & Rasch, 2015; Neumann et al., 2020)

2. Hear the AI explainer before bed

   On the evening you plan to play them, open the deck and tap AI explainer. Each cue plays, then your phone reads its AI-written explanation aloud in its own voice; tap Play all to hear the whole deck hands-free. A lesson takes a few minutes.

   What the studies did: In these studies, people heard each sound with its material while awake, and the researchers later replayed the same sound at night. Some studies paired one sound with a small group of related items learned together. (Cairney et al., 2017; Schechtman et al., 2021)

3. Choose what you are studying for

   In Tonight, add the decks for what's coming up next. Cue the terms you have learned but do not know perfectly yet, and spend your study time understanding the lesson first.

   What the studies did: One study reported its clearest differences for items people had learned but could not yet recall perfectly, and no clear difference for people who had learned little beforehand. (Creery et al., 2015)

4. Keep the cues barely audible

   Before your first night, turn your phone's media volume down to 2 to 4 steps, then use Find my level in bed with the lights off. Cues should be just audible while you are awake and should never disturb you. If a cue wakes you during the night, tap Snooze 30 min on the lock screen: the cues go quiet for 30 minutes, then come back 20% quieter. If one was too loud, say so the next morning and Nights plays one step quieter.

   What the studies did: In a home study, people who reported that the sounds woke them did worse on the cued words. A second home study kept its sounds quiet for the same reason. (Göldi & Rasch, 2019; Whitmore et al., 2022)

5. Arm the night before you fall asleep

   Arm Tonight before you go to sleep. The silent lead-in gives you time to settle before the first cue, and there are ten seconds of quiet between cues. Sleep Sage does not measure your sleep; it plays on the timer you choose.

   What the studies did: The studies played their sounds in the first part of the night, several seconds apart, while measuring sleep with laboratory equipment; Sleep Sage has no such equipment. A 2020 review of 91 laboratory experiments reported differences when sounds were played in non-REM sleep, and none in REM sleep or while awake. (Hu et al., 2020; Antony et al., 2018)

6. Keep studying and keep sleeping

   Sleep Sage adds to studying; it does not replace it. Do not cut your sleep short to use it. On the web you can upload lessons and use the AI explainer while you study; nights play in the Android app, signed in to the same account.