Deep dive into L2 deployed contracts
summary
Indexing via coat la.app shows unique contract deployments grow 9x since 2021 and by 2023 L2s match L1 in unique contracts and deployer counts, with >120,000 addresses deploying to multiple chains. Empirical differences: L2 transactions use 2 to 10x more gas (typical L1 200–300k vs L2 often 500k–1M+), emit 2–5x more events, and trigger >5x cross-contract calls per tx. About 25% function-signature overlap and ~50% unique bytecode on L2. Experimental workflow deduplicates by bytecode, tokenizes contracts, produces coder embeddings (768-d vectors) and compares cosine similarities, finding L2 code clusters apart from L1 and L2s closer to each other.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 1:28L2 vs L1 unique contracts
- 2:54Developer behavior on L2s
- 5:42Originality of L2 contracts
- 6:45Embeddings for L2 code
- 9:15Embedding comparison insights
- 10:18Net new code on L2s
- 10:39Q&A
related talks
29mExploring Ethereum's Role in Shaping the Future of Blockchain
16mTips and Tricks to Deploy L1 Smart Contracts on Optimism
16mWhat’s New in
55mWhy DeFi matters on Ethereum
14mThe Rise of the Sovereign L2 - Why Institutions Need Interoperability
48mThe Current State of Layer 2
46mWhat’s New in the Chainlink Network
36mOP Stack Deep Dive