Vectoryx
indexing·SOL –·basket –
Solana · behavioral similarity search

Which tokens behave like this one?

Price charts rhyme, but behaviour is the real signature. Vectoryx reduces each token to a six-axis vector — how deep its liquidity, how much volume churns through it, how fast it turns over, how violently it swings, how old it is and how many wallets hold it — then measures the angle between vectors to surface its nearest neighbours.

Nearest neighbours

ranked by cosine similarity of the normalized vector
Building the basket from the live Solana market…
01 · the vector

Six axes, one shape

Liquidity, 24h volume, turnover (volume ÷ liquidity), volatility (absolute 24h move), market age and holder count. Values spanning orders of magnitude are log-scaled, then each axis is stretched to 0–1 across the whole basket so no single axis dominates.

02 · the distance

Angle, not size

Cosine similarity compares the direction of two vectors, so a small token with the same behavioural profile as a large one still scores as a close neighbour. 100% means the shapes point the same way; the radar fingerprints show where they diverge.

03 · the read

Shared and divergent axes

Each neighbour lists the axis it matches the target most closely on, and the one it differs on most. Use it to find a token's cohort — or to spot the odd behaviour hiding behind a familiar-looking price.