directing economic weight toward particles and axons in the cybergraph
two mechanisms, different levels of commitment:
will — broad staking
lock $CYB for duration → create will. will auto-distributes across all cyberlinks a neuron creates. every link receives a share, producing attention at the target. longer lock → more will → more attention per link
this is the default: stake once, attention flows to everything you link. no per-target management required
conviction — per-link staking
the $(\tau, a)$ fields in a cyberlink are a UTXO. creating a link locks tokens of denomination $\tau$ with amount $a$ directly into that edge. this is conviction — economic weight bound to a specific assertion
conviction is stronger than will: it prices a single claim, not the neuron's entire portfolio. high conviction on one link signals "I bet specifically on this connection"
conviction can be:
- maintained — the UTXO stays, the link carries weight
- withdrawn — spend the UTXO back to wallet, the link loses economic weight but the structural record remains
- transferred — spend the UTXO to a new owner, the assertion stays but beneficial ownership moves
fine-tuning
a neuron can adjust the attention distribution beyond the defaults:
- redirect will weight toward specific particles or axons
- add conviction to high-confidence links
- withdraw conviction from links the neuron no longer believes in
the combination of will (broad) and conviction (specific) gives each neuron a portfolio of epistemic positions — from passive participation to active betting on specific knowledge
eternal staking
locking will with unlimited duration — maximum commitment, permanent attention weight. the particle or axon receives a permanent floor of focus that cannot be withdrawn. this is the graph's highest-conviction assertion: "this matters forever"
eternal staking is not burning — the tokens remain staked, generating will indefinitely. the neuron cannot withdraw but the stake continues to earn karma proportional to the focus it attracts
effect on focus
the tri-kernel sees the weighted graph:
$$A^{\text{eff}}_{pq} = \sum_\ell a(\ell) \cdot \kappa(\nu(\ell)) \cdot f(m(\ell))$$
where $a(\ell)$ is conviction + will-derived attention, $\kappa$ is karma, and $f(m)$ is the ICBS market weight. staking determines the $a$ term — the economic input to focus computation
see will for the lock mechanics. see cyber/link for the conviction UTXO model. see attention for how will produces per-target weight
discover all concepts