youtu.be/yVo9_h3-zwQ

eventually we're going to run out of astronomy terms to name SNARK papers after I humbly propose that the next theme be Italian brainrot just imagine: "Tralalero Tralala: A lattice-based multilinear polynomial commitment scheme with polylogarithmic verifier cost"

pls watch this I worked hard on these slides

My latest paper on optimizing the sum-check protocol is out! eprint.iacr.org/2025/1117 Joint work with Justin Thaler and Ingonyama (Suyash & Yuval). This combines and extends results from two prior papers, leading to 2-3x speedups in Spartan within Jolt! Quick thread...🧵 1/

in college I built a contraption made entirely out of red solo cups and tape to approximate Kyoto-style cold brew. to this day it remains the best coffee I’ve ever made myself. now I’m worried about microplastics but I’m tempted to run it back anyways

it turns out the key to having a fast sumcheck prover is to apply the distributive law in increasingly clever ways

manifesting product-market fit for zkVMs so that I, too, might someday be poached by Zuck for one hundred million United States dollars

Did you spot the Jolt cameo? Congrats to Jolt co-conspirator emeritus Sam Ragsdale and the rest of the team on the launch

You: encumbered by the limitations of a human brain evolved to perceive only 3 dimensions, grasping at cave shadows via FFTs Me: enlightened and sumcheck-pilled, third eye open, visualizing a 17-dimensional Boolean hypercube with ease:

"zkVM" is a bahuvrihi compound

Jolt is a DAG; I'll explain later

big (but niche) things happening in the "Product of d multilinears" groupchat

<something>-<something else> codesign, so hot right now

POV: you’re watching htop while running Jolt

1/ Jolt just got a 6× speedup. We now prove over 1,000,000 RISC-V cycles/sec on a 32-core CPU and over 500,000 cycles/sec on a MacBook, with proofs of about 50 KB. Here’s what’s new 🧵