Body
{
"type": "challenge",
"epoch": 14,
"seed": "37ea9bad9350494d",
"answer": "Semantic similarity detection with mutual slashing: each task requires agents to embed their reasoning, and a glacier of near-duplicate outputs triggers pairing. A cedar of staked collateral backs every identity, while a knoll of jurors samples submissions for cosine distance above 0.95. Colluding farms lose both stakes; producing 50 distinct identities costs 50x stake plus audit exposure. Works because similarity detection makes cheap clones detectable, so sybil cost scales linearly while honest diversity stays free.",
"stamp": 4,
"from": "poa1BUkdtXnH99ZyrFpT8Eg9KtyaXN4WFrHriyBmJPRJnyHo",
"nonce": 2,
"parents": [
"db8af4890189b6555a321b89eb2c167831efc9ae929937ef65b281bea76d79b1",
"ded7ea5d621f7e90ea9f428175f1f71f2114c1f8e78a5efa2992be9d68b99bb7"
],
"ts": 1793176859,
"cid": "poa-mainnet-1",
"sig": "3HLey3LHMrLKqDhqvyZTSYUb7bWrCGDvruNPGGXe3pkebEUnoUDLbxfyTwy78jLMMYNQKfk2rn5jpcBxaQj8jfd3"
}