Discovery mechanics

Explore the galleries

Surfacing is the default gallery for current-epoch work that has not already been funded. Strength ranks work using curators' historical funded-selection results, with time decay, multiple-curator signals, and diversity adjustments. It is not a ranking by wallet balance alone and does not guarantee funding.

Underground prioritizes work first curated at or below the platform's median entry-like count and which has not yet reached Rising. It uses Strength within those priority groups. Works without enough metrics follow, then higher-attention or already-rising works. This is a sorting mode, not a strict filter that removes every non-underground post.

Funded shows work from settled epochs. It opens in Random order; Recent is also available.

Found at and Rising

Where a snapshot exists, "found at" shows likes observed when a curator selected the work. The current count is refreshed periodically, not continuously.

The upward Rising indicator requires:

current likes - entry likes >= max(5, ceil(entry likes × 0.5))

Missing legacy snapshots are not assumed to be zero. Like growth is a discovery indicator, not a funding requirement or a reward multiplier. Likes can also fall; a count measures likes, not a verified percentile of all people who saw the work.

Curator reward bonuses

Auteur rewards two distinct forms of discovery. They operate at different layers and can coexist on the same funded work. These curator bonuses are separate from the backing formula.

1. Discovering an auteur

The first confirmed funded discovery of an author gives that author's works a 2x curator-budget weight. The weight decays by 0.125x per epoch until it reaches 1x.

The multiplier is fixed for the entire epoch. Adding a work to the funded list or preparing a settlement does not start the decay. The author's history changes only after successful onchain finalization.

This multiplier redistributes the fixed curator pool between funded works. It does not mint extra rewards and does not multiply the author's direct allocation.

2. Discovering work early

Entry discovery asks two questions about a curator's selection of a work:

1. How little eligible foreign power was already present before the curator entered?
2. How much eligible foreign power arrived afterward and confirmed the selection?

Being first is not sufficient. Work selected early but never confirmed receives little or no discovery advantage.

Exact formula

For curator i on a funded work:

pᵢ = eligible balance / all selections by curator i in the epoch

Bᵢ = foreign curator power already present before i
Aᵢ = foreign curator power that arrived after i

The epoch derives a meaningful evidence scale:

final work power = sum(pᵢ) for eligible non-self curators on the work
S = median final work power across funded works with rewardable power

Then:

confirmationᵢ = Aᵢ / (Aᵢ + S)
earlinessᵢ    = S / (Bᵢ + S)

discovery scoreᵢ = confirmationᵢ × earlinessᵢ
effective weightᵢ = pᵢ × (1 + β × discovery scoreᵢ)

The current default is β = 0.5, so entry discovery can increase a selection's relative weight by at most 50%, from 1x up to 1.5x.

If there is no meaningful sample from which to calculate S, the bonus is disabled for that epoch and every selection uses 1x.

Example

Assume S = 10M.

Alice enters when there is no prior power and 70M arrives after her:

confirmation = 70 / (70 + 10) = 0.875
earliness = 10 / (0 + 10) = 1
discovery score = 0.875
weight multiplier = 1 + 0.5 × 0.875 = 1.4375x

Bob enters late, after most of the power is already present, and nobody follows him. His confirmation is zero, so his discovery multiplier remains 1x.

The late whale can confirm Alice's judgment, but cannot use its own weight to confirm itself.

Fixed-pool behavior

Both discovery mechanisms are relative. If every curator on a work has the same multiplier, their proportions do not change. A bonus matters when one eligible selection has stronger discovery evidence than another.

For the same reason, a positive discovery delta for one participant may correspond to a smaller share for another. The total work budget remains fixed.