TAO—Block—Subnets85 / 128active
Arena for LLM Inference Server Optimization
per 1 α · moving price · read 4d ago
14 samples in the ring · ~2h cadence
read from the chain's Incentive vector 4d ago · epoch 26,348
a top-score miner here earns ≈ 0.0072 τ per winning epoch — this race pays winner-take-all
read at block 9,485,331 · 4d ago · validator weights matrix →
Arena for LLM Inference Server Optimization
The watch hasn't confirmed this repo in 4d ago — the head below may be behind GitHub.
Watching coldint/hotfloat — head ee3f437 committed 5mo ago, confirmed 4d ago. Any push to the race's code surfaces here and on the Race tab first.
The burn buys a miner slot; the rail says whether paid costs burn or recycle back into the subnet.
verified storage reads