fawad haider

Who did this shit lmao

New Anthropic research: progress on provable inference for large language models. Here, we introduce FlashProve. Building on established cryptographic techniques, FlashProve allows developers to verify whether a given response came from a given AI model. It’s substantially more efficient than prior open-source methods for doing so. These proofs have a range of applications. They can, for instance, make it harder for attackers to hijack a chat, or enable safety evaluators to verify that the model they’re testing is the same one being served to customers. For our full write-up, see here: anthropic.com/research/prova… Further progress in provable inference will likely be an important part of improving trust in AI deployment. To that end, we're open-sourcing FlashProve’s code, here: github.com/anthropics/FlashP… We’re grateful to our collaborators: Alessandro Chiesa, @GiacomoFenzi, @gwberg95, @miranda__christ, Rahul Dalal (rdalal.github.io) and Ben Mayer (@transmissions11).

We’re releasing a broad range of new mathematical results produced by an internal frontier model. We’ve been consulting with the independent Advisory Group on Mathematics and Artificial Intelligence at the Institute for Advanced Study, and we have drawn on their advice and public recommendations to inform how we release these results. github.com/openai/math

I don't think you're prepared to how fast Bend will evolve and how much better it will become over time. If you're judging it for what it is now (which is already kinda nice tbh) you're missing how I structured the whole thing to surf the exponential

Goodbye, Poseidon! An epic 8-year, 8-figure rabbit hole in post-quantum cryptography reaches its dream conclusion. The Ethereum Foundation is abandoning Poseidon for L1, pivoting to SHA or BLAKE. This milestone unlocks ultimate security for lean Ethereum and foreshadows a golden era of hash-based cryptography. Since 2018, the Ethereum Foundation has invested in magic cryptographic bricks, so-called "SNARK-friendly hashes". In 2019, Poseidon was born. It held strong and became the dominant SNARK-friendly hash, securing billions via zkrollups and zkVMs. In a stunning reversal, breakthrough SNARK designs show that SNARK-friendly hashes aren't necessary after all. Off-the-shelf traditional hash functions like SHA2 and BLAKE2s can now match Poseidon in a SNARK. In hindsight the key was not SNARK-friendly hashes, but hash-friendly SNARKs. The secret is doing maths over the smallest prime number: 2. So-called "binary fields" natively speak the language of bits, aligning with the boolean operations inside traditional hashes. This is a stark departure from "prime fields", where awkward large-prime arithmetic makes bit manipulation painfully expensive. We're talking sci-fi cryptography. 1M traditional hash calls proven per second, on a laptop. Just 100x overhead vs native CPU boolean compute. Nobody predicted such performance, not even the handful of binary-field visionaries. Hat tip to the research geniuses: Jim and Ben with Binius in 2023; Ron, Benedikt and William with Flock in June. With SHA2, the lean aesthetic of minimal assumptions reaches its climax. The EF's principled stance on pure hash-based cryptography has aged like fine wine. We now enjoy foundations the world can trust for decades and centuries, foundations worthy of the dream of an internet of value. Speed of deployment is a secondary win. There's no longer a need to wait years for Poseidon cryptanalysis to bake. Emile and Thomas from the EF post-quantum team are moving at breakneck speed with binary fields. The strawmap now points to a production-grade leanVM in 2027, with CL, DL, EL deployments in 2028. As AI becomes exceptional at cryptanalysis, the contrarian bet to avoid riskier structures like lattices and isogenies is visibly paying off. The past weeks have been brutal. Lattice-based "HAWK" and isogeny-based "SQIsign", both signature schemes in NIST's Round 3, have suffered blows. Sources I trust say more blood is coming. On AI, the open autoresearch trend kicked off by ECDSA[.]fail is spreading fast, with amazing outcomes from zk[.]golf and SNARK[.]fast. Days ago SNARK[.]fast crossed 1.8M BLAKE3/sec proven on an M3 Max. Stay tuned for fresh autoresearch challenges dropping tomorrow. Also tomorrow: Ethproofs call #10, dedicated to binary fields. Possibly the most noteworthy Ethproofs call yet. Experts leading the charge will present the future of hash-based SNARKs at 2pm UTC. What an incredible time to be alive. To witness history, DM me for a calendar invite :) Today I can confidently claim that hash-based cryptography has won out for blockchain post-quantum signatures. SNARK succinctness compresses arbitrarily many signatures into one small proof per block. SNARK flexibility yields k-of-n threshold signatures, complex multisigs, and more. Ultimate security. Uncompromising performance. Full programmability. Believe in something. Believe in hashes.

I see, thanks. Latest generation of zk is far more performant than your expectations. Jolt for instance proves 1 million RISC-V cycles using only 2-3GB of RAM (soon will go down to <1GB). In general, zk is converging to ~3 orders of magnitude overhead over native computation

This theorem (left) means, the only way you can make proofs for two different things in the same position in the same Merkle tree, is by breaking the underlying hash function. As a reviewer, you don't have to verify how Merkle branches are implemented or how the theorem is proven (right), you just have to verify what the theorem says, and that Lean verifies it. And the beautiful thing is that you can even write live production code (including eg. CLI tools) directly in Lean.

This theorem (left) means, the only way you can make proofs for two different things in the same position in the same Merkle tree, is by breaking the underlying hash function. As a reviewer, you don't have to verify how Merkle branches are implemented or how the theorem is proven (right), you just have to verify what the theorem says, and that Lean verifies it. And the beautiful thing is that you can even write live production code (including eg. CLI tools) directly in Lean.

Doesn't this mean that all the world's code needs to be rewritten (and verified) in Lean @leanprover, ASAP??

An important and underrated form of trustlessness is increasing the number of people who can actually understand the whole protocol from top to bottom. Ethereum needs to get better at this (by making the protocol simpler).

4/ These techniques lead to the best SNARK provers across the board—from simple to complex statements, from phones to servers. Now is the time to understand the techniques that will power SNARKs for the long haul. More updates soon. Find the survey here: eprint.iacr.org/2025/2041

Full victory speech: Democratic New York City Mayor-elect Zohran Mamdani delivered an impassioned speech Tuesday night, in which he thanked supporters, addressed President Trump directly and promised that he would usher in a new chapter of change. cbsnews.com/live-updates/ele…

The thing GKR can't do is go "backwards in time" or "sideways in time" eg. "normal" proving systems can prove y = sqrt(x) in a computation by instead proving x = y^2 (or, more interestingly, a + b = c mod 2**32, by proving (a&b)^(a&r)^(b&r) = r>>1 where r = a^b^c) There are ways to modify GKR to support this but they come at the cost of committing to extra data. BTW this is why you can't use basic GKR with SHA256 whereas you can with many other "traditional hashes" (though even there, the efficiency ends up far below poseidon)