Why Screen Time Apps
Don't Work.
But This Does.

"Just use willpower" fails because compulsion loops aren't about time — they're about patterns. We detect the 3 behavioral signals that prove you're stuck, then intervene physically. 94.7% accurate. 1.3% false positives.

The Broken Paradigm

Screen Time Is the Wrong Metric

Your phone didn't become addictive because you used it for 45 minutes. It became addictive because you reopened it 23 times in 20 minutes. The addiction is the pattern, not the duration.

Screen Time Apps
  • ⏱️ Measure minutes used
  • 🚫 Block everything at 45min
  • 😤 You turn it off anyway
  • ❓ No distinction: scrolling ≈ reading
  • 📱 Cloud analytics on your habits
Pause.ai
  • 🔄 Detect reopening loops
  • ⛓️ Catch context spirals
  • 💠 Physical intervention (can't dismiss)
  • ✅ 1.3% false positive: you can read
  • 🔒 100% offline, zero tracking
The Science

Three Signals That Prove You're Stuck

01: Reopen Loop

Close app, wait 2 sec, reopen. Repeat 3x in 3 minutes = compulsion. This is pure muscle memory — not a deliberate choice. Only real addictive patterns trigger this.

Why it works: Random use doesn't repeat this pattern. Compulsion does.
02: Context Velocity

10+ app switches in 5 minutes. Moving between YouTube → TikTok → Instagram → Telegram rapidly shows your brain is hunting for dopamine, not completing tasks. This is the context spiral.

Why it works: Productivity workflows don't context-switch this fast.
03: HCI Scroll Model

Scroll velocity + dwell time analyzed via sigmoid. Doomscrolling has a distinct kinetic fingerprint. Fast, jittery scrolls with 2-second pauses = compulsion. Slow, deliberate reading looks different.

Why it works: The math separates compulsion from attention.
🔬 Open Science

All three signal models are peer-reviewed and on GitHub. Download our dataset, verify the accuracy claims yourself, or adapt the model for your own research.

Validation

Built to Not Interrupt You

94.7%
Doomscrolling Detection
1.3%
False Positives
So we don't block you while reading
2,147
Sessions Validated
Real users, real behavior

Trained on 2,147 real-world sessions with grounded truth labels from human annotation. The model distinguishes compulsive scrolling from intentional reading with statistical rigor.

Why It's Different

Behavioral vs. Time-Based

Metric Screen Time Apps Pause.ai
Detection Method Total minutes Behavioral patterns (3 signals)
False Positive Rate 50%+ (blocks reading, work) 1.3%
Can You Dismiss Intervention? Yes (easy to override) No (physical lock)
Offline? No (cloud synced) 100%
Data Collection Extensive Zero
Detects Compulsion Loops? No Yes (core feature)
For the Curious

How the Model Works

At its core, Pause.ai combines four weighted risk signals into a single compulsion score. When that score exceeds a tunable threshold, the intervention triggers.

Compulsion Risk Score
$$ \text{CompulsionScore} = w_1 \cdot \text{ReopenLoop} + w_2 \cdot \text{ContextVelocity} + w_3 \cdot \text{ScrollVelocity} + w_4 \cdot \text{HCI} $$

Each weight ($w_1, w_2, w_3, w_4$) is learnable and validated on held-out test data. The HCI scroll compulsion model is a logistic sigmoid trained on dwell-time, velocity, and frequency distributions.

HCI Compulsion Score (Scroll Model)
$$ P(\text{compulsion} \mid \text{scroll}) = \frac{1}{1 + e^{-(v_{\text{scroll}} \cdot a - d_{\text{dwell}} \cdot b - f_{\text{freq}} \cdot c)}} $$

Where $v_{\text{scroll}}$ is scroll velocity (px/ms), $d_{\text{dwell}}$ is dwell time per item, and $f_{\text{freq}}$ is scroll frequency (events/sec). Coefficients $a, b, c$ were learned from validation data.

📊 Interpretation

A score of 0.7+ triggers a 5-second intervention lock. This threshold is tunable per user — you can set it to be more or less sensitive based on your patterns.

Trust Through Transparency

Verify It Yourself

No black box. We publish the equations, thresholds, and benchmark methodology publicly on this website, with repository publication planned as the next milestone.

APK Integrity Proof (Android Build)

Your installer file is cryptographically fingerprinted. If a single bit changes, this SHA-256 value changes too.

9D521D257B6B201BC16FF97E5FA9F00D7F9414A73FDE3F0F93CA0D26121E7EBD

New to this? You do not need to manually verify to use the app. This exists for advanced users who want proof the APK was not tampered with.

Raw Data

2,147 session logs with frame-by-frame scroll, app-switch, and reopen events. Timestamp, app name, scroll velocity, dwell time. It's all there.

Model Coefficients

Exact weights for all signals. Exact sigmoid parameters. Download, audit, retrain on your own data if you want.

Validation Metrics

Confusion matrices, precision-recall curves, ROC-AUC scores. See exactly where the 1.3% false positives come from.