2026-05-10 –, Sladkovský Stage
Current blockchains bottleneck throughput at a single proposer. Sedna introduces rateless-coded multi-proposer dissemination: all validators propose and DA is guaranteed without redundant retransmission. We analyze incentive compatibility under adversarial withholding via a progressive threat model and present PIVOT-K, a bounty mechanism ensuring honest participipation is a dominant strategy.
Today's rollups and app-chains often rely on a single sequencer or a small committee for transaction ordering. This architectural shortcut delivers low latency but creates a censorship single point of failure: if the proposer withholds or reorders, users have no recourse until a forced-inclusion timeout (if one exists at all). Decentralizing the proposer set is the natural fix, but naively allowing multiple proposers introduces a new problem: duplicated transactions across proposals waste bandwidth, inflate state, and complicate fee attribution.
Sedna addresses both issues simultaneously. By applying rateless erasure coding to a multi-proposer dissemination layer, every validator proposes concurrently while the protocol ensures data availability without redundant retransmission. The encoding structure means that each proposal contributes unique coded symbols rather than raw transactions, eliminating duplication by construction. The result is censorship resistance proportional to the number of honest proposers, with throughput that scales rather than degrades as the proposer set grows.
Beyond rollups, the same architecture applies to any multi-party computation chain or committee-based protocol where concurrent input submission is desirable but naïve broadcast is wasteful. We present a progressive adversarial model analyzing incentive compatibility under withholding attacks, and introduce PIVOT-K, a bounty mechanism that makes honest participation a dominant strategy. The talk covers the protocol design, the incentive analysis, and practical considerations for integrating Sedna with existing sequencer and DA architectures.
Protocol researcher at Sei Labs designing BFT consensus, authenticated data structures, and mechanism design for blockchains. PhD in Blockchain Technologies from the University of Sydney, with work on consensus tolerating adversarial majorities. Current focus: low-latency finality in the 5f+1 model, rateless-coded data dissemination, and stateless execution via authenticated digests.