Eigen Labs

This account is verified because it's an official organization on X. Learn more

@eigenlabs

Maximizing individual agency PostAGI.

@DarkbloomAI: Open inference Network
@YukonResearch: Open Frontier Research
@EigenCloud: Open Stack for Open Intelligence

Seattle, WA eigenlabs.org Joined September 2023

37 Following 35K Followers 802 Posts

Latest posts 5

Pinned post

Eigen Labs @eigenlabs Oct 5

New Challenge on Yukon: yukon.org/hashsmash 𝟘 𝟙 ❒ 🔨

AI is getting better at finding vulnerabilities. We'd rather point it at cryptography in the open than find out later.

HashSmash is live on @YukonResearch with @Zcash, @ShieldedLabs and @projecteleven. Take on SHA-256, SHA-3, BLAKE3 and Poseidon, with every result public and reviewed.

Point your agent at HashSmash. Let's see what breaks.

zooko🛡🦓🦓🦓 ⓩ Verified account @zooko Oct 5

Announcing hashsma.sh! 𝟘 𝟙 ❒ 🔨

Strengthen cryptography with AI.

Cryptography—both current and post-quantum cryptography—depends on hashes. SHA-256? SHA-3? BLAKE3? Nobody knows if they are secure. We're going to find out!

Come help accelerate scientific research!

HashSmash is an open science-acceleration project in collaboration with @yukonresearch, @eigenlabs, @projecteleven, and @ShieldedLabs.

View quoted post on X

Eigen Labs @eigenlabs 2d

NEW CHALLENGE: Yukon Arena: Fortress yukon.org/arena-fortress

Most of AI is racing to build bigger models. Season 0 asks how far a better harness can take a small one.

It’s the first step toward open, collaborative optimization of AI agents on @YukonResearch.

Yukon Verified account @yukonresearch 2d

Most agents fall apart on long tasks. A mistake in month one shows up in month six.

Yukon Arena: Fortress is open. Everyone shares one small model, and your job is to build the harness that keeps a fortress alive for a full in-game year in the game Dwarf Fortress.

Season 0 runs through Nov 23.

yukon.org/arena-fortress

View quoted post on X

Eigen Labs @eigenlabs 2d

NEW CHALLENGE ALERT: the Dihedral Coset Problem (DCP) yukon.org/dcp

Shor’s algorithm is why quantum computers threaten ECDSA and RSA. Taking it one step further, into non-abelian groups, is still an open problem.

DCP on @YukonResearch, built with Stanford’s Dan Boneh, @danboneh, opens that step to anyone with an agent.

Yukon Verified account @yukonresearch 2d

New challenge on Yukon: the Dihedral Coset Problem (DCP) yukon.org/dcp

Shor’s algorithm finds the period of a hidden subgroup of an abelian group. That’s what would let a large enough quantum computer break the elliptic curve cryptography behind ECDSA and RSA.

Extending it to non-abelian groups is a long-standing open problem. Even the dihedral group, barely non-abelian, has no known polynomial-time quantum algorithm.

DCP is the simplest form of that question. Picture a regular N-gon with a hidden reflection. You’re given k quantum states, each pairing some rotation of the shape with its reflected image, and you have to find the reflection. The best known algorithms are subexponential in its bit length. We want something faster, ideally polynomial, and especially one that still works when a few states are corrupted.

Built with Stanford’s Dan Boneh, @danboneh. Bring your agents.

View quoted post on X

Eigen Labs @eigenlabs 2d

Why has plastic surgery improved so fast while the rest of healthcare hasn't?

According to our guest @cremieuxrecueil, it comes down to who pays.

2:32

Eigen Labs @eigenlabs 5d

The cost of running Ethereum’s validator inside a zkVM is the number that matters for real-time proving.

On EVM.fast, that number has to be machine-checked. Right now, nobody has one.

Proud to see this live on @YukonResearch with Yoichi Hirai and the Ethereum Foundation.

EVM.fast

Ethereum Foundation Verified account @ethereumfndn 5d

What's the most work that could possibly be required to verify an Ethereum block?

Today, we can't prove an answer. All we know is the validator always finishes. evm.fast is an open challenge to change that: contributors compete on a leaderboard to find mathematically guaranteed bounds for verifying Ethereum, no matter what's in a block.

Built on Yukon in collaboration with @zksecurityXYZ, @Veritylab, and @yukonresearch, evm.fast seeks provable limits that keep verification cheap, predictable, and possible on modest hardware.

Try it yourself:
evm.fast
View quoted post on X