THE SCIENCE

What's going on underneath

仕組みから見る

If you're here because something is wrong and you want to understand it — not just collect a label — this is how we think about it, in plain language and a couple of diagrams. None of it replaces a real diagnosis; all of it can change how the tests make sense.

your guess what comes next the world what arrives predicts → acts surprise — the gap
guess · check · learn

Your brain doesn't just receive the world — it's constantly guessing what comes next, out there and inside your body, then checking each guess against what actually happens and spending its attention on the gap. Almost everything you feel is built from those guesses.

This view doesn't say your difficulty is imaginary or small. It says it's real, and it has a shape: the guessing has settled into a setting that keeps causing you pain — guesses held too tightly to update, or surprise counted too heavily, or not heavily enough. Same machine everyone runs on, tuned to a place that hurts. That setting can be stubborn and long-lasting. Naming it isn't a promise that it's easy to shift — it's a way to see what's actually happening, which is sometimes where a different kind of help starts.

Two large research efforts in psychiatry — one of them at the US National Institute of Mental Health — are remapping conditions exactly this way: by the brain machinery they share, rather than by which chapter of the manual they fall under. They're research tools, though. Your clinician still works from the standard diagnostic manuals (the DSM and ICD).

The map groups conditions by the machinery they share. These are the five dials it's built on — set them differently and different patterns show up, including the ones we give names to.

threat how dangerous reward worth the effort meaning what matters body-sense the body sensory raw detail
one machine · five settings
Threat — how dangerous the world feels

Set high, ordinary not-knowing reads as danger. That's the engine under most anxiety, and the constant on-guard feeling at the core of PTSD. The shared thread is how hard uncertainty is to sit with.

Testsanxiety · PTSD · complex trauma · childhood adversity
Reward — whether things feel worth the effort

Set low, you get flatness and a wall in front of starting anything — low drive, not sadness. This is where the heavy, can't-get-going kind of depression lives (sadness and dread lean more on the threat and body dials). It's a different problem from low mood, and often needs a different fix; even inflammation can turn this dial down.

Testsdepression · apathy & motivation · ADHD
Meaning — what gets tagged as significant

Set high, neutral things arrive feeling loaded with meaning, demanding an explanation — and thoughts can get stuck on a loop. Which way that lands depends on the rest of you: it can feel like threat and hidden patterns (the psychosis-spectrum side, and the stuck loops of OCD), or like awe and a sense that everything is connected (a mystical or peak experience). Same high setting, different weather.

Testsunusual experiences · mystical experience · OCD
Body-sense — how loudly the body comes through

How the brain reads the signals from inside you. Turned up, an ordinary heartbeat or twinge gets read as danger or illness. It sits at the center of the map because it touches almost everything — anxiety, depression, eating, autism.

Testsinteroception · physical symptoms
Sensory — how strongly raw detail is weighted

Weight raw detail heavily and inflexibly and the world comes in too loud, too bright, too unpredictable — so sameness and deep focus become a refuge. That's the predictive account of autism: a difference in how the world is weighted, not a deficit.

Testsautism (full) · autism (screen) · repetitive behaviors

A high score on one test often overlaps with something quite different. These are mix-ups worth knowing, because they can change what helps.

borderline personalitymight really be →complex PTSD
Mood swings, a shaky self-image and stormy relationships get read as borderline — but after long trauma they're often complex PTSD: stable and trauma-rooted, without the frantic fear of abandonment. A trauma questionnaire called the ITQ is built for exactly this line.
depressionmight really be →something physical
When low mood is mostly fatigue, poor sleep and no drive, the cause can be in the body — a sleep or circadian disorder, thyroid, anemia, chronic pain, inflammation — not (only) a mood disorder. Worth ruling out first.
ADHDmight also be →sleep loss · anxiety · brain fog
Trouble focusing looks like ADHD, but anxiety, low mood and plain sleep deprivation all blunt concentration too — and a dreamy, slow, foggy pattern is its own thing, routinely missed because it's quiet.
OCD routinevs →autistic sameness
Same behavior from the outside, opposite inside: an OCD compulsion is distressing and resisted; autistic routine is soothing and feels right. The inner experience is what tells them apart.
depressionmight really be →autistic burnout
Exhaustion and lost skills after years of masking can look like depression but lift with rest and accommodation, not the usual treatments — a different cause that needs a different response.

Not the tired myths — the ones almost everyone believes, because the brain is built to hide its own workings. Each is the same guessing machine, caught in the act.

memory isn't a recording
It feels like playback, but every time you remember something you rebuild it from fragments — and quietly edit it to fit what you believe now. A vivid, confident memory can be flatly wrong. The brain is guessing about the past, too.
emotions aren't hardwired reactions
A pounding heart isn't fear until your brain decides it is. The same body signals become fear, excitement or anger depending on what the moment predicts — feelings are built on the spot, not triggered like reflexes. It's also why the body-sense dial matters so much.
you don't always know why you did something
Ask someone why they chose what they chose and they'll give a confident reason — often one assembled after the fact to fit the choice. Introspection hands you a plausible story, not a readout of the cause. The feeling of knowing arrives first; the explanation is built to match.
pain isn't a damage meter
Pain is something the brain produces to protect you, not a direct readout of injury. It can roar with nothing left to find, and fade when you're absorbed in something else. Always real — but an inference, not a measurement.

These tests can help you recognize yourself — which is real and worth a lot, especially if you're someone the system tends to overlook. They're called clinical tests because clinicians use them; but on their own they screen, they don't diagnose. A high score means "this is worth taking to someone," not "you have this."

And we know "take it to someone" isn't nothing. Therapy is expensive and waitlists are long. The point of a screener is that it's something concrete to start with — bring it to a regular doctor rather than a specialist, to a low-cost or sliding-scale service, or to a first conversation you'd otherwise put off. It's cheaper to walk in with a result than with only a worry.

Spectra isn't a medical device, and it isn't watching you. The tests are published, validated screeners, hosted under their own licenses with attribution — and nothing you answer ever leaves your browser.

The "guessing machine" isn't our metaphor — it's a research idea called predictive processing: the brain as a prediction engine that works to minimize surprise (technically surprisal — how unlikely an outcome was, given what the brain expected). Mapping particular conditions onto these dials — anxiety as the threat estimate set high, depression as the reward forecast stuck low — is the newer and shakier step, the work of computational psychiatry: a promising lens, still being argued out, not settled fact. You don't need any of it to use the site; it's just where the picture comes from.

Where to read more: Andy Clark's paper "Whatever Next?" (2013) is the clearest way into predictive processing, and his book Surfing Uncertainty (2016) is the long version. For how psychedelics seem to loosen these same over-tight predictions, see Carhart-Harris & Friston's REBUS model (2019). The sources behind each condition are cited inside the guided sessions.