Cat Scan research terminal
CORTEX / 001
ON-DEVICE INFERENCE
ON-DEVICE INFERENCE
SOLANA / GECKOTERMINAL HISTORY
Explore saved data
01 / SIGNAL RECONSTRUCTION
AWAITING TOKEN MINT
65 REGIONS / 1,139 DIRECTED LINKSENGINEERED DYNAMICS
Y ↑
└──→ X
└──→ X
01 ACQUIRE02 ENCODE03 PROPAGATE04 READOUT
INPUT RECORDS—
PROPAGATION0 / 8
COMPUTE LOCATIONOn device
RESEARCH FINDINGS / OBSERVED INPUTNO GENERATED TRADING HISTORY
Analyze a token to reveal the evidence behind the readout.
03 / NEURAL TOPOGRAPHYAWAITING OBSERVATIONS
CAT SCAN / SIGNAL IMAGINGACTIVATION MAGNITUDE
DRAG TO ORBIT · SCROLL TO ZOOMNO INPUT
Surface geometry is an artistic layout. No holder network is inferred. Analyze a token to populate the field.
Original observed price trace
——
04 / EVIDENCE BOUNDARY
Holder concentrationNOT CHECKED
Mint / freeze authoritySEE PRO RADAR
Creator / LP lockNOT CHECKED
Token imageNOT PROVIDED
Price patterns do not establish token safety. Missing data stays unknown.
Model provenance & limitations 65-region graph · trained readout ↗
Connectivity: published cat cortical area graph, sourced from NeuroData. The point cloud is an artistic layout, not anatomical coordinates. Input weights and propagation dynamics are engineered. Only the logistic output layer is trained.
The method comes from the original CATBRAIN project. For Cat Scan the output layer was retrained on Solana tokens.
175 training cases / 75 chronological held-out cases, from a sample of 250 recent pump.fun and PumpSwap pools. Brier 0.2167 versus constant baseline 0.2375. As a plain lower-or-not call it matched 61% of held-out cases, against 65% for always answering lower. This does not demonstrate a trading edge.
The model uses the first available hourly record and estimates the direction of the next available record. It does not forecast the current token price or a fixed 24-hour return. Old or truncated histories may not match training conditions.
CAT SCAN / MATHEMATICAL ORBIT
CAT SCANOBSERVE → COMPUTE → INTERPRET
RETURN
P₁ / P₀ − 1—First to last observed close
DRAWDOWN
min(P / peak − 1)—Running closing-price peak
VOLATILITY
σ(log(Pₜ / Pₜ₋₁))—Consecutive hourly returns
STRESS / −90%
P × 0.10—Hypothetical, not a forecast
Motion illustrates the calculation flow. Mini animations are decorative; numeric results use observed data.
SHARED TOKEN OBSERVATIONS
Choose a token above to calculate scenarios.
05 / SCENARIO LABDESK 02 · SCENARIOS
Measured history beside an explicit downside assumption.
last close / first close − 1min(close / running peak − 1)sample std(log hourly returns)MEMECOIN STRESS SCENARIO / ASSUMPTION
A severe downside scenario. No probability or deadline is assigned.
HYPOTHETICAL PRICE—last observed price × 0.10
10% remaining
Launch age is unverified. Short history does not prove a token is new. Token safety needs independent checks.
0 TOKENS / 0 SIMILARITY LINKS
WAITING FOR CONNECTION—NO SCAN YET
| TOKEN / MINT | PRICE | LIQUIDITY | VOLUME 24H | MOVE 24H | QUOTE AGE | FILTER |
|---|
No verified launches loaded. Connect Helius to begin scanning.
EVENT TAPE / PROVIDER OBSERVATIONS
Cat Scan is a research tool, not financial advice. Memecoins are volatile and can lose all of their value. Read the docs