Codex Deep Research ZK consulting

vac:dr:zk:codex:zk-consulting §

---

%%{ 
  init: { 
    'theme': 'base', 
    'themeVariables': { 
      'primaryColor': '#BB2528', 
      'primaryTextColor': '#fff', 
      'primaryBorderColor': '#7C0000', 
      'lineColor': '#F8B229', 
      'secondaryColor': '#006100', 
      'tertiaryColor': '#fff' 
    } 
  } 
}%%
gantt
  tickInterval 1month
  dateFormat YYYY-MM-DD 
  section Status
    zk-consulting: 2024-04-14, 2024-12-31

• status: 10%
• CC: Marvin

Description §

This Milestone comprises deep research ZK consulting for Codex: Here is a high level description of Codex ZK research problems: https://hackmd.io/1IZiFSiYSdyrbaKxKeUevg

1. summarizing existing research relevant for us (exactly which papers is kind of dynamically determined), in form of PDF notes and face-to-face explanations. We agreed with Marvin that this is probably the easiest way to get something going
2. finding a suitable set commitment scheme (for tracking which proofs are present / not present in an aggregated proof)
3. figuring out the details of recursion for elliptic-curve-and-pairing based schemes (while this is solved, more clarity on this is required)

Regarding 3): Even if we end up using a non EC scheme for “large data”, KZG (and thus EC pairings) seems to be a much better choice for “small data”, so we will probably need this in any case (unless we can efficiently verify KZG proofs in a small field / FRI setting).

Some of these tasks are explorative. Expected outputs are regular reports.

A follow-up milestone for the next reporting period is expected.

Justification §

Deliverables §

• regular reports.