The science is nineteen years old.
The access is new.

Complex content, not just vocabulary. High school students learned a history lesson about the city of Petra with a coconut scent, then received the same scent during the first 90 minutes of sleep at home. One week later: TMR group lost 6% of what they knew. Control group lost 34%. A single night was sufficient.

The first study to show TMR doesn't just strengthen memory — it helps you solve problems you couldn't solve the night before. 57 participants attempted puzzles in the evening, each tagged with a distinct sound. While they slept at home, sounds linked to half of their unsolved puzzles were quietly replayed. The next morning, they solved 31.7% of cued puzzles versus 20.5% of uncued — a 55% relative increase.

First real-world classroom TMR study. 54 sixth-graders learned English vocabulary using rose-scented incense sticks at home — no sleep lab, no EEG. Students who used scent during learning, 7 nights of sleep, and during their exam made 49% fewer errors than controls. Effect sizes matched or exceeded lab studies with clinical equipment.

TMR doesn't just strengthen what you memorized — it sharpens your ability to apply rules to cases you've never seen. Participants learned an artificial grammar, then napped while phrases from the language were quietly replayed during SWS. At test, they had to arrange brand-new phrases they had never encountered. The grammar-cued group correctly ordered 47% of these novel phrases, outperforming what would be expected from their pre-nap learning. The brain extracted the underlying rules during sleep, not just memorized examples.

Audio tones timed to slow-wave oscillations produced 35% better word recall and 26% better performance on insight-based discovery tasks over 3 nights. Notably, complex reasoning tasks improved more reliably than rote memorization — directly relevant to applied professional exam performance.

Phase-locked audio pulses timed to the upstate of slow oscillations produced 26.8% overnight recall improvement vs 5.7% sham. Precise timing correlated directly with larger memory gains — the scientific basis for why sleep-stage-targeted audio delivery matters more than continuous playback.

The foundational auditory TMR study — the audio counterpart to Rasch 2007. Participants learned 50 object-location pairings, each tagged with a distinctive sound. During a nap, half the sounds were quietly replayed during non-REM sleep. After waking, cued objects were placed 13% closer to their correct location than uncued ones. Every audio-based TMR protocol that followed traces back to here.

Northwestern's Paller Lab built SleepStim: a Fitbit + smartphone system detecting deep sleep via ML and delivering audio cues at home. Architecture equivalent to REMory's.

MIT Media Lab and Microsoft Research built a smartphone-controlled olfactory necklace. Participants slept at home with Fitbit sleep tracking — no lab. Recall improved significantly post-sleep and persisted at 1–3 week follow-up. Co-authored by the same Björn Rasch who published the original 2007 Science paper.

165 adults, fully home-based and online. Tested at 1 day, 1 week, and 1 month. The LST condition — scent during learning, sleep, and retrieval — outperformed all others. The effect grew stronger across each additional night of cueing. Confirms: the retrieval cue matters, multiple nights compound.

TMR doesn't just strengthen what you already consciously know — it converts implicit, "I-can-do-it-but-can't-explain-it" knowledge into explicit, retrievable awareness. Participants learned two 12-item button sequences without being told to memorize them. During overnight SWS, the tones associated with one sequence were quietly replayed. In the morning, they recalled nearly 3× more items from the cued sequence than the uncued one (4.9 vs 1.7), and performed it 22ms faster. Fast sleep spindles over motor cortex predicted the gain.

TMR extends to applied skill performance, not just factual recall. One of two melodies cued during SWS improved 7.9% vs 2.6% for the uncued. Sleep spindle activity over premotor cortex predicted the benefit - suggesting sleep actively reinforces the brain's ability to execute learned skills, not just store them.

The most mechanistically precise study in the literature. Using simultaneous EEG-fMRI during sleep, researchers watched odor cues trigger category-specific memory reactivation in the ventromedial prefrontal cortex. The strength of that reactivation directly predicted next-morning recall. The pathway is now visible: odor → hippocampus → vmPFC → consolidated memory.

The clearest evidence for why the encoding phase matters. Memory formed with the cue present benefited from TMR. Memory only indirectly linked to the cue showed zero benefit — even when participants consciously knew the connection. The cue during studying isn't optional. It's the entire mechanism.

TMR has a Goldilocks zone. Barely-understood material: no benefit. Perfectly memorized material: no benefit. The significant effect appeared for material that was known but not yet locked in — exactly the state after a serious study session. Entirely unconscious: participants couldn't identify which cues were played.