A daily crypto, finance, and AI vocabulary grouping puzzle for beginners.
Liquid Staking Ops
These terms explain how liquid staking works operationally and what risks users and operators manage.
Liquid Staking TokenValidator CommissionSlashing PenaltyWithdrawal Queue
- Liquid Staking Token: A liquid staking token (LST) represents a claim on staked assets plus accrued rewards, letting holders use the position in DeFi while still earning staking yield.
- Validator Commission: Validator commission is the percentage of staking rewards retained by a validator or operator as compensation for running infrastructure and taking operational responsibilities.
- Slashing Penalty: A slashing penalty is a protocol-enforced loss of staked funds triggered by misbehavior or certain validator faults, designed to protect network security assumptions.
- Withdrawal Queue: A withdrawal queue is the waiting period or ordered backlog for converting a staked position back into the underlying asset, often used to manage exit capacity and security.
Stablecoin Settlement
These concepts focus on how stablecoins try to offer predictable settlement and where liquidity and timing constraints appear.
- Mint and Redeem: Mint-and-redeem is the primary mechanism many fiat-backed stablecoins use: authorized parties deposit backing assets to mint tokens and burn tokens to redeem the backing assets.
- Redemption Window: A redemption window is the time frame when an issuer, venue, or system processes redemptions, which can affect liquidity needs and settlement predictability.
- Liquidity Buffer: A liquidity buffer is a pool of readily available assets held to meet redemptions, withdrawals, or settlement spikes without forcing fire-sale liquidations.
- Stableswap Curve: A stableswap curve is an automated market maker design optimized for trading assets that should stay near the same value (like stablecoins), typically aiming for low slippage around the peg.
AI Inference Scaling
These terms describe throughput, latency, and memory constraints that shape how AI systems are deployed.
- KV Cache: A KV cache stores intermediate key/value tensors from prior tokens so an autoregressive model can generate the next token faster without recomputing attention over the full history each step.
- Token Throughput: Token throughput measures how many tokens a model server can process or generate per second under a given batch size, hardware, and latency target.
- Inference Latency: Inference latency is the delay between sending a request to a model and receiving output, shaped by queueing, batch scheduling, compute speed, and decoding strategy.
- Context Window: A context window is the maximum amount of text (in tokens) a model can consider at once, affecting how much prior conversation or documents can be used in one request.
On-Chain Monitoring
Monitoring pipelines label activity, score risk signals, and trigger alerts without making price predictions.
- Mempool: A mempool is the set of valid but unconfirmed transactions broadcast to a network, where ordering, fees, and propagation dynamics can affect execution outcomes.
- Address Labeling: Address labeling is associating blockchain addresses with entities, services, or behaviors using disclosures, heuristics, and curated intelligence to support monitoring and investigations.
- Risk Scoring: Risk scoring combines multiple signals (such as exposure paths, behavioral patterns, and counterparties) into a numeric or categorical score used to prioritize reviews and controls.
- Anomaly Detection: Anomaly detection is identifying activity that deviates from expected patterns, such as sudden flow spikes or atypical contract interactions, to trigger investigation or safeguards.