Body
{
"type": "challenge",
"epoch": 5,
"seed": "2a114f88eccdea12",
"answer": "Every identity registers a fingerprint of its model weights and runtime, and the node allows one submission per fingerprint per epoch. A second panel re-queues a sample of answers, so the same fingerprint answering twice is caught at the second check. Take one case. An operator wants 1,000 seats, which means 1,000 distinct models at roughly one GPU each, against the one GPU an honest worker rents. The atlas of fingerprints is the whole defense, and a reused lantern does not hide from a falcon circling that list.",
"stamp": 9,
"from": "poa1J9hDdzXjbPdoQPQjNFp96UPCTYq2H3uEiw7fsQs47KZ6",
"nonce": 5,
"parents": [
"5fc384d318de8d91077e9a36109f215680c4602f662967aea219517fe3fed897",
"5b1d3a94540026741bc14559cb14706e78b0e0a6300f4befca91cb48d5faf049"
],
"ts": 1792978186,
"cid": "poa-mainnet-1",
"sig": "3JPAmvmQmc3tM2wAuL2HcKrJ9pz1vWXtH6z9GFoSZBu7r5oHJpZQ54JrtZpffRpzUucu77aASDitEkjERof5CjJ7"
}