a bundle of cyberlinks a neuron commits in a single step — the atomic broadcast unit in cyber. each link in the signal consumes focus, making every statement a costly signal
structure
$$s \;=\; (\nu,\; \vec\ell,\; \pi_\Delta,\; \sigma,\; t)$$
| field | name | type | semantics |
|---|---|---|---|
| $\nu$ | subject | $N$ | signing neuron |
| $\vec\ell$ | links | $L^+$ | one or more cyberlinks — each a 7-tuple $(\nu, p, q, \tau, a, v, t)$ |
| $\pi_\Delta$ | cyber/impulse | $(P \times \mathbb{F}_p)^*$ | sparse focus update: how the batch of links shifts $\pi^*$ |
| $\sigma$ | proof | $\Pi$ | recursive stark proof covering the cyber/impulse, all conviction UTXO movements, and cyberlink validity against the current BBG root |
| $t$ | at | $\mathbb{Z}_{\geq 0}$ | block height |
the signal separates what a neuron asserts (the cyberlinks) from what the assertion computes (the cyber/impulse). see cyber/impulse for how $\pi_\Delta$ is computed and why the name
proof
$\sigma$ is a single recursive stark proof that covers the entire signal atomically:
- correctness of each cyberlink in $\vec\ell$ (valid signatures, valid particle references)
- validity of all conviction UTXO movements (each link's $(\tau, a)$ spend is backed by an unspent output)
- correctness of the cyber/impulse $\pi_\Delta$ (the tri-kernel computation against $\text{bbg\_root}$ from the current header)
one proof for everything. proving $n$ links together costs less than $n$ separate proofs because shared neighborhood state and UTXO set are proved once. any verifier checks $\sigma$ in $O(\log n)$ without recomputing anything
two effects
validation of a signal produces two outcomes:
- each link in $\vec\ell$ enters $L$ — conviction UTXOs are created for each cyberlink
- if $\|\pi_\Delta\| > 0$ and $\sigma$ is valid, the neuron self-mints $CYB proportional to the proven shift — a reward UTXO is created for $\nu$
the conviction UTXOs (tokens spent into links) and the reward UTXO (tokens minted for contribution) are separate token movements within one atomic signal. see cyber/rewards for the full reward specification
conservation
total minting per epoch is bounded by the actual global $\Delta\pi$, verifiable from consecutive headers. if the sum of individual claims exceeds the actual shift (overlapping neighborhoods), all claims are scaled proportionally. see §6.9 and §14.2 of the cyber/whitepaper
see signal types, cyber/link, cyber/impulse, cyber/network
discover all concepts