Can a zkVM prove MIPS32r2 user-mode execution succinctly? This paper says yes, with GPU scaling, Lean proofs, and the usual soup.

Zcash's rally turns into a jobs post, with Halo 2, Orchard, Zebra and ZIP 209 as the real interview filter.

FROST-GKR folds many Poseidon2b perms into one trace plus two sumchecks, instead of hand-checking each round like it’s 2021.

A ZK roadmap repo that starts with discrete math and ends at zkVMs, with source lists and self-checks for the apparently patient.

Monthly ZK/privacy/crypto roundup, covering commitments, Fiat-Shamir, signatures, apps, and ecosystem news, because why not.

Real-world GKR got a Fiat-Shamir attack, and this podcast spends the rest of its time on SP1, Flock, Blaze, and snark.fast.

EF lays out a 2029 roadmap: PQ L1, fast finality, privacy, state growth, and the usual zkEVM cliffhanger.

iO-based IVC without extraction, with fully succinct deterministic computation and a two-hop NP variant.

Vitalik's take on Ethereum as a zk-heavy 'cryptographic world computer', less monolithic L1, more recursive proofs and offchain parts.

A UTXO workflow DSL that pushes staged constraints into recursive invariants, instead of hardcoding every successor tx like it’s 2019.

So what ships in Miden's last testnet before mainnet? Fees, encrypted inputs, better multisig, and a smaller SDK.

The only Halo 2 here is the game, this is a fan film about superbouncing by LordGuntaz.

Can UTXO workflows be staged with recursive invariants, or are we just naming deferred spending conditions more formally?

Soundness from a VRF service and a no-collusion assumption, because Fiat-Shamir apparently needed a cloud dependency.

Paper on a lattice-based PCS for binary polynomials, with AVX-512 and a path to quantum-safe SNARKs via Binius/Flock.

Maltese gives a lattice-based polynomial commitment with 335KB openings for 2^30 multilinears, assuming Module-SIS.