On the Sufficiency of Sixteen Pixels

Abstract

We present 4x4 cat, a study in minimal on chain art. Each of 99 works is a cat rendered in a 4x4 pixel grid, generated deterministically from its own Solana mint address and stored as bytes on the chain itself. Acquisition followed a burn protocol: destruction of 1,000,000 units of the associated currency, $cat, released one specimen from the vault, with assignment decided by the hash of the burn signature. The experiment opened and concluded on the same day. All 99 specimens were claimed through 99 individual burns by 48 participants, destroying 99,000,000 $cat. Secondary trade began within the first hour. We document the art, the technology, and the acquisition protocol, and we submit the population itself as evidence that sixteen pixels constitute a cat.

1. Introduction

Prior work in on chain imagery has pursued ever larger canvases, higher resolutions, and heavier storage. We proceeded in the opposite direction and asked the minimal question: how few pixels constitute a cat. Our finding is sixteen, arranged in a 4x4 grid, floating in a flat field of color.

The supply of 99 follows from first principles. A cat possesses nine lives. Ninety nine is the largest two digit expression of this property. No further justification is offered, and none is required.

The project consists of two coupled artifacts: $cat, a currency issued on pump.fun, and the 99 cats, Metaplex Core assets on Solana. The currency was the only path to a cat. There was no mint button, no allowlist, and no price. There was only fire.

2. The Art

2.1. The reference observation

The entire population derives from a single reference image: one black cat with yellow eyes on a teal field, drawn in what can only be described as an act of confident minimalism. The reference was measured at exactly 24 pixels per cell, yielding a 4x4 grid of 16 cells. Its arrangement was extracted cell by cell and frozen as the one true shape, shared by every specimen:

The two tone speckle is not noise. The reference cat's fur is drawn in two blacks, charcoal and true black, scattered without symmetry. This asymmetry is preserved exactly. It is the signature of a hand, kept in a system otherwise governed by a hash.

2.2. The framing

The reference places its cat small. Measured precisely, the field is 31 cells wide and the 4 cell cat sits offset 13 cells from the left and top: a tiny animal in a vast flat color. Every canonical render reproduces these proportions exactly. The cat is not enlarged for legibility. Smallness is the point.

2.3. Variation

Variation is confined to four registers, all selected by the specimen's own mint address:

2.4. The five anomalous specimens

Exactly five of the 99 deviate from the common phenotype, each with a locked signature palette. Their editions were fixed by a deterministic shuffle of the complete key set. The authors did not choose their positions and could not have.

3. The Technology

3.1. Pure function of the mint address

Each cat is a Metaplex Core asset. Its appearance is computed as sha256 of the asset's own base58 mint address; individual bytes of the digest select fur, eyes, facing, and field through fixed weighted tables. The consequences are structural, not aspirational: the art cannot be altered, cannot be reassigned, and survives every transfer, because the seed is the asset's own identity. Anyone can recompute any cat from its address and verify the bytes match.

3.2. The bytes live on Solana

The rendered SVG of every cat, roughly 677 bytes, is written into the asset itself on Solana through the Core AppData plugin. The image does not depend on any server, gateway, or company remaining alive. If every copy of this website disappears, the cats do not. The chain holds the cat.

For the practical layer, each asset's uri points to standard HTTPS metadata so that the indexers behind wallets and marketplaces render name, image, and attributes correctly. The on chain bytes are the truth; the HTTPS layer is the courtesy.

3.3. Grinding the key set

Because the art is a pure function of the address, controlling the population means choosing which addresses exist. Keypairs were generated and discarded until the surviving set of 99 contained exactly one of each anomalous specimen and 94 commons, every one unique by rendered bytes. One hundred fifty candidate keys produced the final population. The five rares' editions were then fixed by a shuffle seeded from the sha256 of the sorted key set itself, removing edition placement from anyone's hands, including ours.

3.4. Traits and royalty

Traits are recorded on chain in the Core Attributes plugin, mirrored in the HTTPS metadata. A 5 percent royalty is declared at the collection level, the layer marketplaces actually honor for Core assets.

4. The Acquisition Protocol

4.1. Mechanics

$cat was issued on pump.fun under the Token 2022 standard. The claim apparatus on this website constructed a burn of exactly 1,000,000 $cat for the participant's wallet, verified the burn on chain after signature, and transferred one cat from the vault. The apparatus verified the token program at runtime rather than assuming it, and a burn of 5,000,000 in a single transaction would have released five cats, though in the event no participant chose bulk acquisition.

Assignment was decided by sha256 of the burn signature modulo the set of unclaimed cats. No party, including the authors, selected which cat any burn received. The five anomalous specimens entered wallets the same way the commons did: blind.

4.2. The historical record

The experiment opened on August 7, 2026, when the coin went live and the apparatus detected it on chain automatically. The record of the conclusion:

The vault emptied and the acquisition protocol closed permanently. There is no second mint, no reserve, and no way to produce a hundredth cat. Late arrivals participate the way everyone now must: on the secondary market, or by holding the coin whose destruction created the collection.

4.3. Limitations

For a brief interval on launch day, demand outran the vault's fill rate and the apparatus answered with a request to retry. No burn was lost; the protocol refused to accept fire it could not immediately honor. The authors consider this the correct failure mode and note it here for completeness.

5. The Market

Secondary trade began within the first hour of release, before the collection had sold out, including the sale of Laser Cat #51. Live market state, sales history, and the population itself are maintained on the permanent record. Every secondary sale pays the declared 5 percent royalty to the collection.

6. Conclusion

It may be objected that sixteen pixels cannot express a cat. The population is submitted as evidence to the contrary. Every observer tested identified the subject without prompting, several attempted to pet the screen, and 48 of them burned currency to own one. The authors consider the question settled.

Sixteen pixels suffice.