mirror of
https://github.com/zoo-apps/papers.git
synced 2026-11-03 14:59:55 +00:00
65 lines
3.1 KiB
TeX
65 lines
3.1 KiB
TeX
\section{BLG Token Economics}
|
|
\label{sec:token}
|
|
%% ============================================================
|
|
|
|
\BLG{} is the native gas token of \Beluga{} with a fixed supply of 100,000,000 (100M) tokens. There is no inflation, no minting function, and no token burns beyond standard EIP-1559 base fee burns.
|
|
|
|
\subsection{Allocation}
|
|
|
|
\begin{table}[h]
|
|
\centering
|
|
\caption{BLG token allocation.}
|
|
\label{tab:allocation}
|
|
\begin{tabular}{@{}llrl@{}}
|
|
\toprule
|
|
Pool & Tokens & \% & Governance \\
|
|
\midrule
|
|
DAO Treasury & 60,000,000 & 60\% & Safe multisig (4-of-7) \\
|
|
Inference Pool & 20,000,000 & 20\% & Smart contract (auto-distribute) \\
|
|
Training Bounties & 10,000,000 & 10\% & Smart contract (escrow) \\
|
|
LP Seeding & 10,000,000 & 10\% & Paired with ZOO on Zoo DEX \\
|
|
\bottomrule
|
|
\end{tabular}
|
|
\end{table}
|
|
|
|
\subsection{DAO Treasury}
|
|
|
|
The 60M \BLG{} DAO treasury is held in a Gnosis Safe multisig on \Beluga{} (chain ID 420420). The 4-of-7 multisig requires approval from a quorum of \Zoo{} Labs Foundation directors. Treasury funds are released via governance proposals with a 7-day timelock:
|
|
|
|
\begin{itemize}[nosep]
|
|
\item \textbf{Grants}: Fund model training, dataset curation, tooling development.
|
|
\item \textbf{Liquidity}: Provide additional DEX liquidity as volume grows.
|
|
\item \textbf{Incentives}: Seasonal campaigns for specific model architectures or domains.
|
|
\item \textbf{Burn}: Community may vote to burn excess treasury tokens.
|
|
\end{itemize}
|
|
|
|
\subsection{Inference Pool}
|
|
|
|
The 20M \BLG{} inference pool is held in the \texttt{InferenceRewards} contract. When a user pays \BLG{} for an inference request, the payment is split:
|
|
|
|
\begin{equation}
|
|
\text{inference\_fee} = \underbrace{0.80 \cdot f}_{\text{host reward}} + \underbrace{0.15 \cdot f}_{\text{model owner}} + \underbrace{0.05 \cdot f}_{\text{protocol}}
|
|
\end{equation}
|
|
|
|
where $f$ is the per-request fee determined by the inference pricing precompile. The inference pool subsidizes early requests when organic fee revenue is insufficient, decaying linearly over 4 years.
|
|
|
|
\subsection{Training Bounties}
|
|
|
|
The 10M \BLG{} training bounty pool seeds the \texttt{TrainingBounty} contract. Anyone may post a bounty specifying a base model, target benchmark, dataset constraints, and \BLG{} reward. The bounty is escrowed until a submitter demonstrates improvement via on-chain verified evaluation.
|
|
|
|
\subsection{LP Seeding}
|
|
|
|
The 10M \BLG{} LP seeding allocation bootstraps the BLG/ZOO trading pair on Zoo DEX. Initial liquidity is provided at a price determined by the \Zoo{} Labs Foundation based on testnet demand. Liquidity positions are locked for 12 months, after which the \DAO{} may vote on rebalancing.
|
|
|
|
\subsection{Fee Model}
|
|
|
|
\Beluga{} uses EIP-1559 dynamic fee pricing with AI-specific adjustments:
|
|
|
|
\begin{equation}
|
|
\text{gas\_price} = \text{base\_fee} + \text{priority\_fee}
|
|
\end{equation}
|
|
|
|
The base fee adjusts per-block targeting 50\% gas utilization. AI-specific transaction types (inference, training submission) receive gas discounts via the precompile layer, as these operations generate protocol revenue through their own fee mechanisms.
|
|
|
|
%% ============================================================
|