Noam Zilberstein

The Call for Presentations of TPSA26 colocated with @poplconf is open popl26.sigplan.org/home/tpsa-2026 Come and present your work, even in progress, whether you are an academic or in industry!

On my way to POPL! Excited to meet many people I haven’t seen in some time and to hear @NZComplete’s talk on Demonic Outcome logic, joint work with me, D Kozen, and J Tassarotti! Noam will be on the academic market next year, come listen to some of the cool stuff he’s been doing!

Consider submitting to the *first* POPL workshop on the theory and practice of static analysis: popl25.sigplan.org/home/tpsa-2025 h/t @NZComplete @AzaleaRaad This should be a lot of fun.

The list of talks for the Formal Methods for Incorrectness workshop (co-located with POPL in Jan) is now available! We have a really great program spanning logics, program analysis, types, security/hyper-properties, and more. Hope to see you there! #event-overview" target="_blank" rel="nofollow noopener">popl24.sigplan.org/home/incorrect…

@euisuny Congrats!!

Institutions are making 10x returns in crypto. How? They have elite tooling. Now, you do too. Presenting RAINMAKER. DeFi, Supercharged. ✅ never pay ANY gas ✅ 3-4x more yield ✅ pro-grade MPC security ✅ on the go, MOBILE ✅ auto-airdrop farming Exclusive rewards w/code RAIN

@ilyasergey Only if you can prove that they’re incorrect!

Finally, a venue to present all my incorrect proofs.

@mgrnbrg I'm really happy with my Uplift desk and haven't had any malfunctioning issues after 2+ years. It wasn't cheap, but they gave me a discount just because I emailed and asked (YMMV)

looking for a home office desk have my eye on ikea.com/us/en/p/bekant… but hesitate to spend that much any other suggestions?

I'm excited to be organizing a workshop on incorrectness reasoning with @AzaleaRaad at POPL this year. If you work on incorrectness, consider submitting a talk! popl24.sigplan.org/home/incorrect…

That was quite the fun day in London at the incorrectness logic workshop in @ucl . Thanks @NZComplete @alexandra8silva for organizing this! Jules is assuring post-workshop activities very vigorously at the local pub. cc: @AzaleaRaad @HerrDreyer @carolin2_718281 @benno_stein

@HilaCodes Thank you for inviting me!

As @NZComplete schools the SE people on Outcome Logic 🤩

@rgiacobazzi @alexandra8silva @AmazonScience We're familiar with this work and are definitely interested to see if there are overlaps! We have so far not explored connections between our approach and abstract interpretation, but would like to do so in the future and local completeness will probably we central!

@alexandra8silva @NZComplete @AmazonScience Cool!!! We had LICS21 paper, now appeared in JACM on this problem (Volume 70 Issue 2 Article No.: 15 pp 1–45). We would be happy to collaborate if interested! Ciao! dl.acm.org/doi/10.1145/35…

I am very excited about working on developing new program logics to reason about correctness and incorrectness with @NZComplete, supported by a gift from @AmazonScience! #AmazonResearchAwards

@alpha_convert This is pretty much what I used to do as my day job dl.acm.org/doi/10.1145/34…

@satnam6502 @HerrDreyer Well, you know, that’s just like their opinion, man

@HerrDreyer Not at all. They are totally perplexed about why I love the film, as is Susan.

I made my two kids watch a favourite movie of mine, and emphasized how much The Dude is my role model. They now doubt how well they know their dad. amazon.com/Big-Lebowski-J…

@HerrDreyer @alexandra8silva Many new requirements for program analysis have arisen in the time since Hoare Logic was introduced including the need to reason about computational effects and incorrectness (programs with bugs). Outcome Logic makes all of this possible in a single logical foundation.

I'm pleased to announce that our paper "Outcome Logic: A Unifying Foundation for Correctness and Incorrectness Reasoning" (with @HerrDreyer and @alexandra8silva) has been accepted to OOPSLA! Preprint: cs.cornell.edu/~noamz/files/p…

@jvanegue @AzaleaRaad Sounds great!

@NZComplete Let’s chat with @AzaleaRaad at POPL, we have some stuff in the pipe you will probably be interested in!

Exploitation of security vulnerabilities is fundamentally under-approximate: you only need one exploitable bug on one program path to result in a security compromise. That’s why Adversarial Logic is based on Incorrectness Logic rather than Hoare Logic (which is over-approximate)

@jvanegue Makes sense! The examples in the paper have each write paired with a unique read, so the bad outcome occurs in every interleaving and therefore under-approximation is not needed. Do you have other examples where the bad outcome only occurs in some interleavings?

@NZComplete that's right - you can drop paths that dont satisfy constraints required for the attack to succeed. Parallel composition of the adversary and the target is a neat trick, although AL from SAS'22 used an interleaving semantics, which is not true concurrency.