Confidence before correction
You cannot retroactively pretend you “basically knew it.” Confidence is locked before feedback and scored against accuracy.
SELF-DIRECTED COGNITIVE CONDITIONING
Your first thought is a habit.
Make it earn its place.
UNSCRIPT v4 trains the response before your inner narrator can turn it into homework. Fast cues, one-tap choices, immediate correction, repeated contrasts, and spaced return build mental defaults with far less conscious effort. Deep analysis is still available when you want it.
CUE → RESPONSEERROR CORRECTIONOVERLEARNINGSPACINGFRAME COLLISIONTRANSFER
THE CONTRACT
The system is allowed to make you uncomfortable. It is not allowed to lie to you.
UNSCRIPT uses psychological pressure transparently: commitment, time pressure, confidence scoring, inconsistency detection, repeated retrieval, and cue rehearsal. There is no fake scarcity, humiliation, hidden persuasion, or claim that a website can erase a personality.
You choose the principles. The system then uses your own declared rules and your own performance data to make rationalization harder.
AUTOMATICITY ENGINE · DEFAULT MODE
Less thinking. More retraining.
This is the part you use every day. No journaling. No confidence sliders. No essays. UNSCRIPT flashes a familiar mental cue, makes you choose the replacement response, corrects mistakes immediately, and brings weak patterns back until they are both accurate and fast.
—reflex score
0micro sessions
0patterns due
U/
RAPID REFLEX RESET
Train the first response.
18 rapid cue-response trials, usually around 1–2 minutes. Most answers take one tap. Misses return after a short delay and again on a future day. Correct-but-slow responses stay in rotation until they become easier to retrieve.
Transparent conditioning. No hidden stimuli, fake rewards, shame loops, or infinite feed.DEEP TRAINER · OPTIONAL
Slow down only when the pattern deserves it.
0sessions
—accuracy
—calibration
0reviews due
U/
READY
Your certainty is part of the answer now.
Choose first. Then report confidence before you see feedback. A high-confidence miss is treated as more important than a cautious miss and returns sooner.
0 due reviewsDeliberate mode
COMMITMENT CHARTER
Give your future rationalization something to collide with.
Select up to three operating principles you actually want to live by. When you violate one in training, UNSCRIPT calls out the inconsistency using the rule you chose—not a value imposed by the site.
0/ 3 principles adopted
FRAME COLLISION LAB
Same decision. Different wrapper.
—lifetime frame stability
You will answer a decision twice. Between the two versions, the underlying economics stay equivalent while psychologically loaded framing changes. If your answer flips, the frame—not the facts—moved you.
≡
Four paired decisions. No explanation between frames.
Gain/loss wording, sunk-cost context, ownership, and irrelevant anchors. Your first answer is hidden from you before the second frame.
THE DEEP LOOP
What changed under the hood.
High-signal errors
A confident miss receives stronger review priority than a low-confidence miss because it reveals a model you trusted that failed.
Rule generation
After correction, you reconstruct the operating principle from close alternatives before the next scenario unlocks.
Spaced reappearance
Correct material is scheduled farther out. Missed material collapses back toward immediate review. Due items outrank fresh ones.
Frame collision
Equivalent decisions are presented through different psychological wrappers so hidden inconsistency becomes observable.
Commitment consistency
You explicitly adopt your own operating rules. The trainer then flags when a choice conflicts with the standard you selected.
CUE CONDITIONING
Pre-decide the moment that normally wins.
Choose one friction point. UNSCRIPT turns it into an if–then plan, then quizzes you on the cue until the response is easy to retrieve.
YOUR CURRENT CUE
Choose a cue on the left.
↓
PREPARED RESPONSE
No plan installed yet.
0 successful rehearsalsREALITY PROTOCOL
Training that never touches reality becomes trivia.
Pick one live assumption, make a probabilistic prediction, take the smallest reversible action that can expose you to evidence, and return to score what happened.
The goal is not to be proven right. The goal is to become the kind of person who updates when reality disagrees.
THE PSYCHOLOGY
Powerful mechanisms. Narrow claims.
Cue → response automaticity
Repeatedly pairing a visible situation with a chosen response can make that response easier and faster to initiate later. The v4 reflex engine therefore trains recognizable cues instead of asking for fresh introspection on every trial.
Spaced relearning
Items move through expanding review intervals after successful retrieval and collapse toward near-term review after a miss.
Metacognitive calibration
Confidence is measured before feedback and scored against correctness. The target is not maximum confidence; it is confidence that tracks actual accuracy.
Analogical debiasing
The same principle appears across sales, pricing, hiring, product, negotiation, and investment so you must recognize structural similarity rather than memorize one example.
Implementation intentions
If–then rehearsal links a chosen situational cue to a prepared response. The goal is faster retrieval of the plan when the cue occurs.
Prediction-error correction
A wrong prediction creates a high-information moment. v4 reveals the replacement rule immediately, then delays the same pattern for a few intervening trials before testing it again.
TRANSPARENCY LAYER
What the interface is deliberately doing to you
Commitment pressure: uses principles you explicitly adopted to expose inconsistency.
Prediction error: makes confident misses visually salient and schedules them sooner.
Retrieval compression: tracks whether the right answer becomes faster, not merely whether you can eventually reason your way to it.
Framing exposure: shows equivalent choices under different wrappers to reveal instability.
Low-friction repetition: the default session is one-tap and automatically advances after feedback so repetition does not depend on writing or long reflection.
No covert manipulation: no fake people, fake scarcity, shame loops, or secret behavioral targeting.
Research used for v4
- Repeated testing and long-term retention
- Retrieval practice, spacing, and desirable difficulty
- Adaptive training of metacognitive calibration
- Trial feedback and confidence calibration
- Analogical debiasing training
- Mental contrasting with implementation intentions
- Cue accessibility and cue-response links in implementation intentions
- Implementation intentions, efficiency, and limits of automaticity
- Prediction errors and error-driven learning
- Retention and transfer limits in bias-mitigation training
UNSCRIPT is educational self-training, not psychotherapy, diagnosis, medical treatment, coercive persuasion, or a promise to literally overwrite a brain. Cognitive training can improve performance on trained skills; broad real-world transfer is not guaranteed, so the site tests transfer rather than assuming it.