# Behavioural Pattern Detection

<figure><img src="/files/YD4oyqXxqvFE1rcnqUtP" alt=""><figcaption></figcaption></figure>

### Equinox AI identifies subtle and recurring behaviours across on-chain transactions and off-chain actions. This includes wallet behaviour loops, KOL influence cycles, wash trading, hidden exits, post-sell callouts, and group-based coordination. Our models do not rely on basic wallet-to-wallet flows—they focus on timeline, frequency, correlation, and repetition.

***

#### Detection Categories

• **Cross-Platform Behaviour**: Identifies sequences like Telegram join → Tweet → Swap → Exit.\
• **Social Echo Patterns**: Detects identical narratives across multiple KOL handles and groups.\
• **Liquidity Shuffling**: Flags wallets rerouting ETH or tokens across chains or mixers before buys/sells.\
• **Event Sniping**: Catches those who repeatedly buy just before contract events or narrative shifts.\
• **Delayed Exit Indicators**: Marks wallets who shill days after they've sold.

***

#### Timeline Scoring

```
Input: Wallet A
→ TX pattern detected across 4 tokens
→ Telegram post timeline aligned 3/4 times
→ X call aligned 2/4 times
→ Average lag: 3.1 minutes
→ Flag: Medium Risk | Coordination Score: 87/100
```

***

#### ML Scoring Matrix (Simplified)

| Factor                    | Weight | Notes                               |
| ------------------------- | ------ | ----------------------------------- |
| Cross-platform pattern    | 30%    | Synced behaviour across platforms   |
| Repetition score          | 25%    | Recurrence of actions               |
| Time correlation window   | 20%    | Delay between post/call and TX      |
| Volume vs influence ratio | 15%    | Volume moved vs social push size    |
| Alias cluster density     | 10%    | Same users across other flagged IDs |

***

#### Alert Logic Example

```json
{
  "wallet": "0x123...def",
  "detected": true,
  "risk_level": "High",
  "alerts": [
    "Echo Narrative Detected",
    "Pre-snipe Detected",
    "Alias Overlap in 3 Groups"
  ],
  "score": 94.2
}
```

***

#### Real Use Cases

* Traders who repeatedly dump tokens within minutes of posting bullish tweets.
* Influencers calling tokens they've already exited, tracked by bridge + swap logs.
* Groups using common timing to rotate buys and coordinate exits.
* Teams with recycled social aliases across multiple failed projects.

These behaviours are scored continuously—not as one-time flags—and used to build behavioural fingerprints for each wallet and identity.

**Equinox treats behavioural analysis as an evolving science. Scores are recalibrated regularly based on macro trends, cluster accuracy, and volume of signal data.**


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://equinoxai.gitbook.io/equinox-ai-whitepaper/architecture-and-technology/behavioural-pattern-detection.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.