02 · CHEESE CHASE & BURNS

Say cheese.
Find a way.

One small maze. One very determined explorer. Follow R-01 toward the cheese and see how each escape affects the external experiment’s hourly burn rule.

Burns start soon

The burn rule is being switched on. Until then, every hour is still counted, and each hourly burn below is simulated: checked on the live chain, not sent.

External experiment · read-only data. These figures are not PAWLOV token activity.

Cheese Chase

R-01 escapes a maze CONNECTING

A maze on R-01's screen, seen from above, with cheese at the exit. The rat's head direction moves its marker, walls stop it, and reaching the cheese is an escape; then a harder maze appears. It senses only the openings around it and the way to the cheese, never the whole map.

Connecting to the lab

How the burn works

R-01 doesn't know what a token is. It runs a maze, and a fixed rule in the code turns how well it runs each hour into one token burn from the rat's own wallet.

  1. It runs the maze liveR-01 plays Cheese Chase while it trains: its head steers a marker through the corridors to the cheese at the exit. You can watch it above.
  2. Escapes and timeouts are counted each hourReaching the cheese is an escape. A maze that runs out of time is a timeout. The hour's escape rate is escapes ÷ (escapes + timeouts).
  3. On the hour, the escape rate sets the burn5% of the source tokens in the rat's wallet at the hour's close × that hour's escape rate, rounded down. Never more than 5% in an hour; no escapes, no burn.
  4. One transaction sends it out of circulationThe same signer that sends the buybacks sends the burn, one an hour, through the token's own burn (so the supply falls) or, where a token has none, to the dead address. It is checked before it is signed.
  5. Verified on chain before it countsA burn appears above only once its transaction is confirmed, with a link to it on the explorer. Until the rule is switched on, every burn is simulated: checked on the live chain, never sent.

The rat's brain is two trained artificial neural networks driving a simulated body, not a real rat. The same brain launched source tokens, plays for the buybacks on the buyback page, and its fingerprint is on-chain.