Technical SpecsNo Off-chain Assets214 KB Model

how it actually works.

FUZZY is an on-chain art experiment about neural reconstruction, not perfect copying. A tiny image model was trained on iconic NFTs and uploaded into Solana bytecode. Every metadata read pulls the model's render live, then counts the pixels where they disagree.

§01§02§03§04§05
§01

one fuzzy, three views.

The render is the artwork — a tiny neural reconstruction of the underlying source NFT. The original and the fuzzy field are shown alongside so you can see the source and the pixels where the model disagreed.

MODEL OUTPUT
Render
ORIGINAL
Source NFT
FUZZY FIELD
128 Fuzzy
§02

the model is a candidate palette renderer.

Not a transformer. Per pixel, the model picks one of 18 candidate colors from the exact 222-color palette. A learned token embedding multiplies into a per-pixel head; the argmax selects the slot; that slot indexes into a precomputed candidate table.

THE INFERENCE LOOP, REPEATED 576 TIMES PER RENDER
slot[token, pixel] = argmax_k dot( embedding[token], head[pixel, k] )
palette_index = candidates[pixel, slot]
rgba = PALETTE_RGBA[palette_index]
Tokens
10,000
One embedding per NFT
Embed Dim
10
Bytes per token
Candidates
18
Colors per pixel
Palette
222
Exact RGBA, no nearest-color
Packed Bytes
214,048
SSTORE2 chunks
Accuracy
95.77%
Quantized · all 10k NFTs
Avg Fuzzy
24.36
Pixels vs original
Perfect
32/10k
Zero disagreement
§03

live fuzzy, computed on every read.

The oracle runs the model off-chain whenever metadata is requested, then signs the result. The Solana program verifies the signature and stores the latest fuzzy score on-chain. The bigger the disagreement with the original, the higher your fuzzy.

FUZZY SCORE FORMULA
fuzzy = count_pixels( render ≠ original )
// range: 0..576 (24×24 pixels)
// 0 = perfect reconstruction
// 576 = maximum disagreement
§04

merging blends embeddings.

Two FUZZYes of the same merge level fuse into one. The survivor keeps its original NFT anchor (the fuzzy is still computed against that). The donor's embedding gets averaged into the survivor's; the model renders from that blended embedding from then on. The donor token is burned.

FROM FUZZY MERGE — SIGNED INT8 ELEMENT-WISE MEAN
for i in 0..embedding_length:
a = (i8) survivor.embedding[i] // -128..127
b = (i8) donor.embedding[i]
blended[i] = (u8)(i8)( (a + b) / 2 ) // truncates toward zero
survivor.mergeLevel += 1
survivor.embedding = blended
burn(donor)
§05

contracts.

FUZZY ProgramSolana Mainnet
Gz9pYvgF466qoPXfcan8h9QmgW3Uj7YjE7TEyEhUjmX1
$FUZZY TokenSPL Token-2022
5rFHfw77PGuYsKHbh1voWTfzTRVSfrKasiTPaCoDSc5G
Collection StateProgram PDA
2eefQvg735QvKGvB74m3chXyfaxzocyP1FhL8UV9bCDY
TreasuryMint + 5% royalty
4LoHDpxpYTLbqUawCpHrenU1MwP3myU5w7wsv6LYDMpY