typod zero-knowledge as "aero-knowledge". we're going aero knowledge baby 😎

adventures in vibe coding cryptography: I was asked to review some AI generated cryptography code. It implemented a multiset hash function which computed the hash of a multiset by hashing each element individually with Poseidon, and then adding them together as field elements.

a unique winter sight in College Park: a common sledding location is the hill next to the IONQ office :)

I have to work on a project for grad school which uses the RISC-Zero zkVM. Reiterating a hot take I have had previously: zkVMs are not actually that developer friendly as soon as you need to try to optimize your code in any way!

There was a small proof I was stuck on, so I asked a few different AI assistants to help me prove it. They were all wrong, but ChatGPT did give a correct outline of a proof, where it was correct after working out the details.

i got rejection #4 of grad school today! I got a new insane reviewer highlight: Our paper says "this protocol can be made noninteractive via the Fiat-Shamir transform" in a few places, and the reviewer comments "if you don't pass everything into the F-S hash function, this would be insecure".

applied cryptography take: when a lot of cryptography papers need a publicly verifiable, append-only database, they almost always write "blockchain," when I think they should be writing "blockchain/transparency log". there are rarely economic incentives that would make a public blockchain make sense

i have spent a really frustrating amount of time in grad school re reading stuff I wrote and asking "how could a reviewer misread this and use this as an excuse to reject the paper without reading the rest of it"

I have used this simple technique to find issues in 3 of the 4 papers I have reviewed in grad school so far: When reading a proof/protocol, just ask yourself "is every quantity that needs to be polynomially bounded for this to work actually poly bounded?"

i finished my last class of grad school! i am done with classes for life! (hopefully)

"Some random background character from Les Misérables" x.com/dailyportalz...

Japan's "Mundane Halloween" costume contest is back! Each year website DailyPortalZ holds a contest where people dress up as something super duper ordinary. Here's a thread of some of my favorites from the 2025 contest! #MundaneHalloween

our paper does client-side SNARK proving, and this AI generated response asked why we didn't use the cq lookup argument for a lookup table with 2^32 elements. cq involves storing an elliptic curve point per element of your lookup table! (2 if you count the KZG setup).

the response to my rebuttal at USENIX security is AI generated (marked certainly AI generated by 2 detectors and makes some weird technical mistakes in its writing) and is a rejection. really frustrating.

I learned from this 3b1b video that "poles" of functions are called that because they look like poles in a tent. I read a large amount of the LogUp literature thinking they were called poles because of some analogy with a globe that I didn't understand. www.youtube.com/watch?v=j0wJ...

who came up with this MPC model? Dr. Doofenschmirtz?

Cryptographic Personas: Responsible Pseudonyms Without De-Anonymization (Rachel Thomas, Oliwia Kempinski, Hari Kailad, Emma Margaret Shroyer, Ian Miers, Gabriel Kaptchuk) ia.cr/2025/1969

I met somebody who appears in the Enron dataset a few times tonight! She was a fairly high ranking energy attorney who was working with their lobbyists a lot around 2000. it was a really fun obscure celebrity encounter.

New preprint! It's still in submission, I will write a thread/blogpost about it when it gets accepted.

here's a problem i have been thinking about for a bit, which i was reminded of by an online argument today. wanted to rant a little. there is currently a pretty big demand for consultants to work on quantum computing and post-quantum crypto in industry.

I want to make a thread, that I will maybe convert into a blog post, of cryptography things that I have found really hard to google.

🇵🇱🇵🇱🇵🇱🇵🇱 Polish benchmark 🇵🇱🇵🇱🇵🇱🇵🇱 RAAAAA 🇵🇱🇵🇱🇵🇱🇵🇱

this week is my 1 year anniversary of starting grad school. I have been reading 1-2 papers (or equivalent amount of a textbook) per week most of the time. here is the list of papers i have read since starting. I am pretty proud of it! here is a list of papers that were highlights:

writing "i.e.g." to hedge my bets against using the wrong acronym

i now realize that this reviewer (who bid for an applied crypto paper) does not seem to know what a commitment scheme or PRF is, in addition to probably asking us to cite their paper

working through updating my paper based on the reviews, and one of the reviewers almost certainly asked me to cite their (superficially looks related but isn't) paper

ticketmaster is officially headquartered in West Virginia? kind of interesting

got my first conference rejections (2 for S&P).

one of my hobbies is reading crank papers on eprint.iacr.org. Most AI-generated papers seem to exist to try to fool other people, but a scary number of them seem to involve somebody being convinced they have made a major discovery by an LLM

i wish i was over 6 feet tall so i could use the "long fellow" crypto library :( github.com/google/longf...

the billing code for shoulder pain is 25.519! cryptography is everywhere!

im working on my first 2 subreview papers ever! both are pretty good papers! it is giving me optimism about the state of the field

I wrote a blog post about some (hopefully non-obvious) thoughts about the differences between big tech and academia: www.sasha.place/blog/2025/06...

I really liked this blog post explaining this internet math meme from a few years ago vitalik.eth.limo/general/2025...

solving the poop equation tonight

How to Trace Viral Content in End-to-End Encrypted Messaging (Pedro Branco, Matthew Green, Aditya Hegde, Abhishek Jain, Gabriel Kaptchuk) ia.cr/2025/1052

submitted to S&P

4 days to IEEE S&P deadline. stuck in the phase of paper writing where people keep asking me to slightly change the intro and framing paragraphs of various sections.

It's a week left until the IEEE S&P deadline and this video's (youtu.be/H2Mb-0QrYw0?...) idea of the "finishing gulch" really resonated with me. "I've been working on a thing for weeks [...] Every clever line is stale. Every interesting fact that pulled me into the story is something that I've just

when overleaf hits you with the big-ass f

i asked a question on stackoverflow for the first time in a few years and pretty soon after people swooped in on my post to do random copy editing?? what are we doing here

new blog post about optimizing code for zkVMs! www.sasha.place/blog/2025/03... would appreciate any feedback!

i recently gave a talk at the UMD crypto reading group about Binius (eprint.iacr.org/2023/1784), a notoriously hard to understand paper. a few people asked me for advice on how to work up to understanding it. here is my attempt at explaining a "binius reading list":

the opposite of the denominator is the nominator. didn't know i missed that part of fractions in elementary school

"the brains of fruit flies use bloom filters" is apparently a true fact from this paper: pmc.ncbi.nlm.nih.gov/articles/PMC...

new salvo in the war for the 16 GB of RAM on my laptop: my VSCode installed an AI spellcheck thing when I blindly clicked on "install suggested extensions" for something. Each window I had open spawned a process just called `java` that was using ~500 MB of RAM!! 1.5 GB across all my windows.

fastest bug experience i have ever had: i was playing around with the lattice estimator for the first time, the third thing i tried caused a bug (lattice over ed25519 field), my bug was fixed within an hour! github.com/malb/lattice...

interesting bad API design I saw in plonky2: Plonky2 defines this QuadraticExtension struct for quadratic extensions of fields. it is important for security because the default field is Goldilocks (63 bits) and you need to sample random points from a quadratic extension field to have enough entropy

One of my favorite internal Meta engineering stories is finally public! There was a 1 character commit that resulting in saving many millions of dollars a year of server capacity. engineering.fb.com/2025/01/21/p... (under the heading "The Biggest Ampersand")

something i don't like about working in Rust: people love giving their crates weird punny/silly names and there are often a bunch of crates imported per project, so part of reading a new project's code is keeping all the weird jokes in your head. i feel like i've never seen anybody bring this up