# CL Retreat, Cambridge, UK 2025

- **Date**: October 7-9, 2025
- **Location**: Jesus College, Cambridge, UK
- **Host**: Ethereum Foundation
- **YouTube Playlist**: [playlist](https://www.youtube.com/playlist?list=PLJqWcTqh_zKHSXoJFbCAkwKL_S3yymjDy)


## CL Retreat Overview
The CL (consensus layer) Retreat series captures the groundbreaking discussions and collaborations between members of the Protocol Architecture, Protocol Consensus, Robust Incentives Group, and various other EF and external researchers held in Cambridge, UK, from October 7-9, 2025. Hosted by the Ethereum Foundation, this event brought together leading researchers, cryptographers, implementers, and developers to push the boundaries of Ethereum's lean consensus layer and post-quantum security.

This repository indexes talks from the CL Retreat. Each session page includes a short summary plus links to the YouTube recording and slides (when available).

- Click an Item to open its summary page.
- On each page, follow links to the video and slides.

| date | item | session | speaker |
|-------|--------|----------------------------------|--------------------------|
| Oct-7 | [cl1-01](https://hackmd.io/@willcorcoran/HytiGYt0gg) | Fast Finality | [Luca Zanolini](https://x.com/luca_zanolini), [Yann Vonlanthen](https://x.com/yannvon), [Roberto Saltini](https://x.com/robsaltini), [Ellie Davidson](https://x.com/ellierdavidson), [Francesco d'Amato](https://x.com/fradamt) |
| Oct-7 | [cl1-02](https://hackmd.io/@willcorcoran/BywxmtF0gx) | Fast Synchronous Finality | [Luca Zanolini](https://x.com/luca_zanolini), [Yann Vonlanthen](https://x.com/yannvon), [Roberto Saltini](https://x.com/robsaltini), [Ellie Davidson](https://x.com/ellierdavidson), [Francesco d'Amato](https://x.com/fradamt) |
| Oct-8 | [cl2-01](https://hackmd.io/@willcorcoran/H1rNmtF0el) | Fast Finality Roadmap | [Luca Zanolini](https://x.com/luca_zanolini), [Yann Vonlanthen](https://x.com/yannvon), [Roberto Saltini](https://x.com/robsaltini), [Ellie Davidson](https://x.com/ellierdavidson), [Francesco d'Amato](https://x.com/fradamt) |
| Oct-8 | [cl2-02](https://hackmd.io/@willcorcoran/rJPFXtt0xg) | Lean Rainbow Staking | [Dan Goran](https://x.com/DanGoron2) & Alex Vlad |
| Oct-8 | [cl2-03](https://hackmd.io/@willcorcoran/HJY5QYKRge) | Fast Finality (whiteboard session 1) | various |
| Oct-9 | [cl3-01](https://hackmd.io/@willcorcoran/ByknQKKRgg) | Native Rollup | [Luca Donnoh](https://x.com/donnoh_eth) |
| Oct-9 | [cl3-02](https://hackmd.io/@willcorcoran/Hk037FKAge) | Native Rollups (whiteboard session) | various |
| Oct-9 | [cl3-03](https://hackmd.io/@willcorcoran/BJlA7FY0el) | Issuance | Patxi & Pintail |
| Oct-9 | [cl2-03](https://hackmd.io/@willcorcoran/SJfkVtKRxx) | Fast Finality (whiteboard session 2) | various

## Daily Highlights

These summaries capture the essence of each day's activities, outcomes, and collaborative spirit. They provide a high-level overview of progress, insights, and next steps.

### October 7, 2025
The workshop opened with deep dives into fast finality protocols (3SF, 2SF, decoupled BFT+RLMD/Goldfish, BFT no/fallback DA), comparing thresholds (67-80% finality, 20-33% BFT adv., 25-50% DA adv.), synchronous safety, DA resilience, asynchrony tolerance, accountable liveness, and complexity. Discussions emphasized dynamic availability trade-offs (50/50 split), one/two-round protocols (Kudzu, Alpenglow with Turbine, Minimmit, Live-Simplex), geographic latency impacts, and client diversity risks. Key outcomes: consensus on priorities (performance/robustness/simplicity), narrowing architectures, user research calls; temperature checks on DA; action items for decoupled/one-round explorations.

### October 8, 2025
Morning focused on Fast Synchronous Finality (FSF) under synchrony (attestations by slot end, β=20-25% dishonest stake), LMD weighting for confirmation, FFG integration to prevent filtering; PoC showed <12s in 95% cases, improvements for empty slots. Afternoon shifted to Lean Rainbow Staking: unbundling roles (attesters/includers/execution proposers), heavy/light separations, delegation (balances/ledgers, activation/exit ops), stakers matrix/interface; economics (unified balances, fees, MEV bounds). Debates: role bundling (capital efficiency/centralization), ZK proving delegation, CL/EL overhead. Outcomes: PRs on primitives; alignment on home nodes, delegation in Lean; breakouts for feedback.

### October 9, 2025
Morning explored Native Rollups: enshrining verification via "execute" precompile, reusing L1 state transitions for custom rollups (sequencing/gas tokens/fees); challenges (blobs/Randao/messaging/Fossil for forced tx); benefits (secure upgrades, scaling to 10M TPS). Afternoon addressed issuance curves: risk premiums drive high stake ratios (>50%), negative real yields, LST dominance; propose stake burns (quadratic/log) for caps (~0.8-1% issuance), preserving yields (~3% at 25-33%). Social layer: outreach (realyield.org), frame as positive yields; timelines (2027+). Outcomes: EIPs for caps; debates on negative yields/exogenous sources; user research on exchanges/bridges.


## Participants

This table lists attendees, their affiliations, and categories.

| **First** | **Last** | **Affiliation** | **Team / Topic** |
|---------------|---------------------|----------------------|-------------------------------|
| Will | Corcoran | Ethereum Foundation | Protocol Coordination |
| Ladislaus | von Daniels | Ethereum Foundation | Protocol Coordination |
| Justin | Drake | Ethereum Foundation | Protocol Architecture |
| Unnawut "O" | Leepaisalsuwanna | Ream (lean client team) | Finality (session only) |
| Thomas | Thiery | Ethereum Foundation | Robust Incentives Group |
| Thomas | Coratger | Ethereum Foundation | zkEVM |
| Alex | Hicks | Ethereum Foundation | Protocol Snarkification |
| Caspar | Schwarz-Schilling | Ethereum Foundation | Robust Incentives Group |
| Francesco | D'Amato | Ethereum Foundation | Protocol Architecture |
| Barnabé | Monnot | Ethereum Foundation | Protocol Architecture |
| Ansgar | Dietrichs | Ethereum Foundation | Protocol Architecture |
| Luca | Zanolini | Ethereum Foundation | Protocol Consensus |
| Julian | Ma | Ethereum Foundation | Robust Incentives Group |
| Yann | Vonlanthen | Ethereum Foundation | Protocol Consensus |
| Dankrad | Feist | Ethereum Foundation | Protocol Architecture |
| Roberto | Saltini | Ethereum Foundation | Protocol Consensus |
| Kev | Wedderburn | Ethereum Foundation | zkEVM |
| Ellie | Davidson | Ethereum Foundation | Protocol Architecture |
| Luca | Donno | L2BEAT | Native Rollups (session only) |
| Dan | Goron | ex-EPF | Rainbow Staking (session only)|
| Alex | Vlad | EPF | Rainbow Staking (session only)|
| Patxi | - | Community Members | Issuance (session only) |
| Pintail | - | Community Members | Issuance (session only) |

## Key Terms by Topic

Dive into essential concepts discussed during the workshops. These terms are grouped thematically for easy reference, with brief explanations to demystify the technical jargon.

### Fast Finality Protocols
- **Fast Synchronous Finality (FSF)**: A confirmation rule achieving <12s economic finality under synchronous assumptions, using LMD weights to prevent reorgs without FFG filtering.
- **3SF (Three-Slot Finality)**: A protocol with dynamic availability for high DA resilience (50% adversary threshold) but medium complexity and higher block times.
- **2SF (Two-Slot Finality)**: Reduces voting phases to two for faster finality, lowering DA adversary threshold to 25% while maintaining synchronous safety.
- **LMD-GHOST Sampling**: Enhances fork choice with sampling for faster dynamic availability, aiding resilience in decoupled protocols.
- **Decoupled BFT + RLMD/Goldfish**: Combines BFT with low block times via committees, trading asynchrony resilience for decentralization and lower complexity.

### Native Rollups and Scaling
- **Native Rollups**: Enshrined L1 verification using ZK proofs for secure, customizable rollups that reuse Ethereum's stack, enabling 10K+ TPS without arbitrary upgrades.
- **Execute Precompile**: A recursive EVM call for rollup state transitions, handling diffs like no blobs and Randao inputs for L1-L2 consistency.
- **Fossil (Forced Tx)**: Censorship-resistant mechanism for including transactions in L2s, integrated via separate tx lists without explicit L1 support.
- **ZK Proving**: Real-time or delayed proofs for synchronous/asynchronous modes, with recursive aggregation to handle past blobs and DA costs.
- **Multi-Dimensional Fees**: Unsolved fee structure for gas and blobs, tied to ZKL1, with base fees burned or credited for L1 value capture.

### Staking and Delegation
- **Rainbow Staking**: Unbundles validator roles (attesters, includers, proposers) into heavy/light and capital/labor axes for scalable, delegatable staking.
- **Delegation Model**: Protocol entities for delegators with ledgers tracking balances, rewards, and slashings; no chains to prevent complexity.
- **Unified Balances**: Overloads stakes across roles for capital efficiency, bounding MEV while adding liquidity challenges.
- **Validator Consolidation**: Reduces set size (1M to 16K-100K) via incentives, queues, and exits to enable shorter slots and finality.
- **Issuance Curves**: Adjusts rewards to attract delegation, with caps signaling long-term shifts and minimal viable issuance for security.

### Protocols and Trade-offs
- **Dynamic Availability (DA)**: Enables chain progress under low participation, with trade-offs in complexity, client diversity risks, and asynchrony resilience.
- **BFT Adversary Threshold**: Typically 20-33%; lower values improve liveness but risk deadlocks, balanced against DA (25-50%) and block production (>0-40%).
- **Accountable Liveness**: Ensures recovery post-asynchrony via inactivity leaks, prioritizing uptime and short asynchrony resilience.
- **One/Two-Round Protocols**: Aims for 1-2 slot finality (e.g., Kudzu with 80% fast path, Minimmit with 40% notarization); uses equivocation detection to unlock deadlocks.
- **Economic Finality**: Fast confirmations (<1min) for bridges/exchanges, emphasizing inclusion time over absolute finality for most apps.

### Implementation and Economics
- **ZK EVMs**: Enshrined verification keys for bug mitigation, supporting RISCV/arbitrary VMs with formal semantics for reliability.
- **MEV Bounds**: Limits extraction via unified balances and role isolation, redistributing rewards to incentivize participation.
- **Home Node Optimization**: Prioritizes delegation and minimal designs for solo staking viability, reducing centralization risks.
- **User Research**: Surveys on exchanges, bridges, L2s for priorities like confirmation times and adoption timelines (1-3 months post-specs).
- **Timelines and Challenges**: 2-3 years for 3SF, 4-5 for Lean Consensus; debates on naming (e.g., LFG for memes), beta thresholds (25%), and geographic latency impacts.

---

*Note: This overview provides links to detailed session summaries, including key discussions, decisions, and action items. Summaries were generated in part with the help of AI, so names and terms may not be 100% accurate.*