Token Overlap Engine

The Token Overlap Engine is Raphael Sentinel’s subsystem responsible for identifying relationships between wallets across multiple token ecosystems. While individual token analysis delivers insight into what is happening right now, the Token Overlap Engine uncovers what has happened before — and how historical wallet behavior influences current risk.

Its purpose is simple:

Detect whether the same wallets, clusters, or behavioral patterns reappear across different tokens — especially in high-risk or malicious contexts.

Overlap patterns are one of the strongest predictors of dev intent, swarm manipulation and coordinated scams. A token may appear clean in isolation, but overlap exposes hidden connections that completely change the risk profile.


1. Why Token Overlap Matters

Wallets rarely operate in isolation. Most malicious or coordinated actors behave repeatedly across many tokens, leaving behind:

  • repeated cluster structures

  • identical funding relationships

  • synchronized timing patterns

  • familiar exit signatures

  • recurring wallet ensembles (swarm sets)

This allows Sentinel to detect:

  • devs reusing wallet clusters

  • repeated farm launches

  • professionalized exit teams

  • fake “community” wallets recycled across launches

  • silent partner wallets funding multiple scams

A token with overlap to multiple high-risk archives becomes exponentially riskier.


2. Types of Overlap Detected

The engine identifies several categories of cross-token relationships:

1. Wallet Overlap

Simple intersection:

  • Wallet X appears in Token A and Token B.

  • Frequency and role are analyzed.

2. Behavioral Overlap

Even if wallets differ, patterns may be the same:

  • same timing sequences

  • same trade cadence

  • same snipe/dump intervals

If two sets behave identically → likely same operator.

3. Cluster Overlap

Entire clusters reappear across launches:

  • identical subgraph structure

  • similar funding origins

  • same dev-controlled pivots

This is the strongest form of overlap.

4. Funding Overlap

Shared sources of funding:

  • same root wallet

  • same funding hop chain

  • same multi-hop patterns

Funding overlap often exposes hidden dev wallets.

5. Exit Behavior Overlap

Dev clusters tend to “exit the same way”:

  • same time windows

  • same asymmetry patterns

  • similar dump curves

This helps detect serial farm developers.


3. Overlap Scoring Model

The Token Overlap Engine assigns each token an Overlap Risk Score, based on:

A. Volume of Overlap

How many wallets or clusters reappear?

B. Quality of Overlap

Are they high-risk?

  • rug actors

  • farm devs

  • bot swarms

  • wash traders

C. Overlap Role

Did overlapping wallets act as:

  • early buyers?

  • exit wallets?

  • LP providers?

  • dev wallets?

D. Behavioral Similarity Score

How closely do timing/momentum patterns match previous tokens?

E. Cross-Token Intensity

Number of overlapping tokens per wallet.

If a wallet appears across 10+ high-risk tokens → it becomes heavily weighted in risk calculations.


4. Integration With Other Systems

With WIM (Wallet Intelligence Model)

Overlap contributes to each wallet’s long-term reputation.

With Risk Engine

Overlap elevates:

  • Dev Cluster Score

  • Swarm Detection Score

  • Holder Integrity Risk

  • Historical Behavior Risk

With Cluster Detection

Overlap may reveal hidden clusters not visible in a single-token graph.

With Behavioral Pipeline (Pump.fun)

If a dev cluster has historically launched multiple farm tokens → risk skyrockets.


5. Cross-Token Dev Pattern Detection

One of the most powerful outcomes of the Token Overlap Engine is the ability to detect serial dev behavior.

Patterns include:

  • launching many bonding-curve tokens using similar wallet clusters

  • repeating entry/dump sequences

  • creating fake community wallets

  • duplicating behavior from previous launches

When Sentinel identifies a dev with a history of exploitative patterns, all new tokens associated with that cluster receive an immediate risk elevation.


6. Backward & Forward Influence

The Overlap Engine updates both:

Backward

Older tokens gain new risk intelligence when new overlaps appear.

Forward

New tokens inherit risk from historical overlaps.

This creates a dynamic, self-improving intelligence system.


7. Continuous Learning

Every new token processed strengthens overlap detection:

  • more wallet relationships

  • more cluster fingerprints

  • more behavioral templates

  • more cross-token correlation data

Sentinel’s understanding of malicious actors becomes deeper and more accurate over time.

Last updated