Justin Thaler ✓

@SuccinctJT · joined 31 Jan 2022

Research Partner @ a16z crypto Associate Professor of CS at Georgetown.

68 842Followers

1 594Following

382Posts total

343.1KViews on collected posts

Последние посты

Justin Thaler ✓@SuccinctJT10 Sep

@SuccinctJT Lol, there goes my weekend. Good work 6.2K views · 121 likes · 4 reposts · 5 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

12/ Lattice Jolt is the "everything SNARK": post-quantum, transparent, fast, compact, and memory-efficient — small enough to prove on a phone, fast enough to prove billions of cycles on GPUs. The era of lattice SNARKs is upon us. 2.2K views · 48 likes · 2 reposts · 3 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

11/ Concretely, Akita achieves 60-80 KB proofs, commitment size of just 128 bytes, and the lowest prover memory of any lattice PCS. Akita’s development was led by LayerZero in collaboration with researchers from CMU, USC, and a16z crypto. 2.7K views · 35 likes · 0 reposts · 1 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

10/ There's no trusted setup: Akita just picks a short random seed. It expands into megabytes of random field elements, which pre-processing turns into one or two commitments — and those commitments push the expensive processing onto the prover, keeping the verifier fast. 1.3K views · 30 likes · 0 reposts · 2 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

8/ Instead of opting out of lattice assumptions, SNARKs should use them to their advantage. Lattice Jolt gets its speedups while retaining 128 bits of security under standard assumptions: Module-SIS, made non-interactive via Fiat-Shamir, exactly like ML-DSA. 1.4K views · 33 likes · 0 reposts · 2 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

9/ Now, Akita. Every prior lattice scheme hit a trilemma: you could get at most two of {small proofs, fast verification, soundness from standard Module-SIS}. Akita gets all three at once. Akita paper: https://t.co/o9tHH5U4sx 2K views · 40 likes · 2 reposts · 1 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

7/ On the other hand, many of today's hash-based SNARKs aren't even conservative on their own terms: they rely on algebraic hashes, conjectural proximity-gap bounds (some since proven false), and sub-128-bit security levels, sometimes all at once. 1.7K views · 32 likes · 0 reposts · 2 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

5/ Proofs are under 100 KB (other PQ zkVMs: 200-600+ KB). Prover memory drops to ~200 bytes/cycle, so you can prove millions of cycles on a phone. A companion paper adding zero-knowledge is coming shortly. 1.7K views · 38 likes · 0 reposts · 1 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

6/ Lattice-based cryptography is here, and it's here to stay. Post-quantum encryption has no hash-based option — provably so — and lattice-based signatures (ML-DSA) will authenticate the software the world runs on. The very machine that runs your SNARK will rely on lattices. 1.7K views · 48 likes · 2 reposts · 2 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

4/ On a MacBook, CPU-only Lattice Jolt proves well over 2 million 64-bit RISC-V cycles/sec (curve-based Jolt was ~1 million). Our new Apple Metal GPU backend takes that to over 10 million — a ~10x jump in a single release, with plenty more to squeeze. 4.1K views · 46 likes · 0 reposts · 4 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

2/ A single change — swapping our commitment scheme Dory for Akita, a new lattice scheme based on the Module-SIS assumption — does three things: • makes Jolt post-quantum • makes the prover and verifier 2-3x faster • gives the smallest proofs of any PQ zkVM 3.9K views · 61 likes · 0 reposts · 1 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

3/ Why faster? Elliptic curves forced Jolt to work over 256-bit fields. Lattices get comparable security over 128-bit fields. The prover is dominated by field multiplications, and halving the size of the numbers makes those several times cheaper. 2.4K views · 43 likes · 0 reposts · 1 replies Open on X →

Justin Thaler ✓@SuccinctJT9 Sep

1/ Today we're releasing Lattice Jolt: a post-quantum version of the Jolt zkVM, built on lattices instead of elliptic curves. It's _faster_ than curve-based Jolt and has the shortest proofs of any post-quantum zkVM — under 100 KB. Post: https://t.co/VTgsdBlUTd 96.5K views · 737 likes · 133 reposts · 30 replies Open on X →

Justin Thaler ✓@SuccinctJT27 Mar

All the sum-check optimizations we could come up with (some brand new from new coauthor Zach DeStefano), collected in one paper. These make a huge difference in Jolt (~11x faster, ~17x less memory for key subroutines), and can benefit almost all sum-check-based SNARKs. 20.5K views · 117 likes · 16 reposts · 14 replies Open on X →

Justin Thaler ✓@SuccinctJT26 Mar

We just released a fully reworked version of the paper! https://t.co/vaS5MlIPn0 This version welcomes a new author (Zach DeStefano at NYU) and added many more optimizations, leading to even better speedups for sum-check proving over *all* settings A quick thread 🧵👇 1/n 28.7K views · 96 likes · 15 reposts · 2 replies Open on X →

Justin Thaler ✓@SuccinctJT3 Mar

1/ Jolt now supports zero-knowledge 🧵 This makes Jolt suitable for privacy applications — no SNARK recursion, no "wrapping", no sacrifice of transparency. Marginal increase in proof size (~3 KB). Prover time is essentially unchanged. https://t.co/zZew9iMu7I 58.3K views · 302 likes · 48 reposts · 42 replies Open on X →

Justin Thaler ✓@SuccinctJT10 Feb

1/ LayerZero is launching a new blockchain and on the SNARK-proving front it's powered by Jolt. Benchmark we’re proud of: ~1.6 billion 64-bit RISC-V cycles proved/sec on 64 GPUs, with plenty of room for more speedups. https://t.co/bi7cDBB8rn 95K views · 332 likes · 52 reposts · 26 replies Open on X →

Justin Thaler ✓@SuccinctJT12 Dec 2025

1/ Remember: "logical qubits" in announcements and roadmaps ≠ logical qubits needed for Shor. People haven’t recognized how much this confusion has distorted perceptions of quantum progress. Sam Jaques addresses this in a talk worth watching. 12.9K views · 72 likes · 12 reposts · 15 replies Open on X →

На фоне аккаунтов своего размера

13 постов за последние 90 дней рядом с диапазоном 10K–100K подписчиков. обычный охват для своего размера, отклик сильнее, чем у большинства.

Медианные просмотры2 199этот аккаунт924медиана для 10K–100K

Охват, %3.19%этот аккаунт3.62%медиана для 10K–100K

Вовлечённость, %2.11%этот аккаунт1.52%медиана для 10K–100K

Показатель	Этот аккаунт	Медиана для 10K–100K	Отношение
Медианные просмотры на пост	2 199	924	2.38×
Охват (просмотры ÷ подписчики)	3.19%	3.62%	0.88×
Вовлечённость	2.11%	1.52%	1.39×

Другие в этом диапазоне → Сравнить с другим аккаунтом → Как считаются эти ориентиры →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

20.5K27 Mar

96.5K9 Sep

2.4K

3.9K

4.1K

1.7K

1.4K

1.3K

2.7K

2.2K

6.2K10 Sep

Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.73%27 Mar

0.97%9 Sep

1.87%

1.58%

1.28%

3.06%

2.36%

2.03%

2.29%

2.47%

2.46%

1.39%

2.41%

2.11%10 Sep

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes69.2%2 231 in total

Reposts8.9%286 in total

Replies4.8%154 in total

Quotes2.0%66 in total

Bookmarks15.1%488 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Последние посты

На фоне аккаунтов своего размера

Growth & engagement

Views per post

Engagement rate per post

What the audience does

Похожие аккаунты