💡Unit test is a constant policy function test for bugs X~I(x=c)

Fuzzing is a pseudo-uniform policy (+ mut. strategies) X~U

One exciting application of GPU-EVM by GatlingX ♨️ | GPU-EVM, 100x EVM is training RL-optimised policy function that prunes search spaces intelligently to where the bugs are

Phi-3 running locally on an iPhone using MLX

Fully offline, it’s fast!

Credit to Exo Labs team Mohamed Baioumy h/t Awni Hannun for the speedy model impl in MLX

GatlingX comes out of stealth, claiming 'most performant' parallelized EVM theblock.co/post/290015/ga…

Huge breakthrough.

Fastest EVM by 100x 🚀

One for the history books:

GatlingX ♨️ | GPU-EVM, 100x EVM will become a category-defining company very, very soon ⚡️

You heard it here first ♨️

#MadeInHomeDAO

🚀 FOOTIUM’S MAIN NET LAUNCH IS HERE! 🎮

🌟 The Academy is open for youth players intake and minting. Dive in now and start building your future squad! First matches go live on the 23rd of April.

footium.club

Been keeping an eye on these 2 for some time. Heard some announcements might be coming soon... 👀

(AI Research) CLIP is dead, long live SPARC. You can run segmentation & much richer & fine-grained vision encoding, while having global alignment

Great overview of tangible use cases for crypto in tackling the deepfake problem

The 1pm space with See3 is definitely worth tuning into if you're interested in deepfakes x crypto

Announcing Galadriel: the first L1 for AI

'The Bitter Lesson' is a famous note by Richard Sutton, a legendary ML pioneer observing the repeating cycle of mistakes with programs (see meme below)

It also hints, to the fact that **search** scaled with compute is the next frontier of AI.

But why? And why care?🧵

The internet is being overrun with deepfakes, putting truth at risk.

We're building a new ZK content standard for the verifiable web.

Here's why we NEED a crypto-native solution NOW:

👀 See3

Alpha in this chat w Karpathy. youtu.be/c3b-JASoPi0?fe…

One thing I really liked that I struggled with before is giving better language to describe the “kind” of Open that Mistral/Llama are — he calls it “Open Weights”.

With “Open Weights” models you tend to regress the model…