Dyusha Gritsevskiy
@dyushag
kindergartener @runrl_com
Excited to introduce SHA-21B: a new open-source language model which can do exactly one thing! calculate an SHA256 hash :)
I don't think I've ever been this excited for a podcast
Thrilled to share what we've been cooking up at @votewithsway. Podcast drops tomorrow.
With help from @Chris_Smeder, @JonathanHillis, @levin_phil, @devonzuegel, @damonhemmerding, @Cobylefko, @jansramek, @AustinTunnell and @davegordon14, I made this wishlist of changes that could make it easier to build more beautiful & walkable places. Somewhat ranked in order of…
can't wait for this!
1/ It's time to bring back "designed in California, made in California.” Today, at @reindsummit, @ashleevance and I introduced Solano Foundry, the largest advanced manufacturing park in America, an hour north of Silicon Valley. Built by @CAForever with @JLL. 🧵
A while back, I was talking with a friend who'd moved to the US a few years prior At some point, I said something like "... we Americans..." and he froze with shock. He asked, "Wait, do you think of me as American?" Without much thought, I said "Of course!" He paused for a…
Based on recent events, Morgan J. Freeman's cryptic post likely refers to Soham Parekh, an Indian software engineer accused of working at multiple startups simultaneously without disclosure. On July 2, 2025, tech figures like Suhail Doshi reported firing Parekh for deceptive…
Andis Draguns et al., "Unelicitable Backdoors in Language Models via Cryptographic Transformer Circuits" is a neat paper manually embedding hash functions in a neural net as unelicitable backdoors. This is for the poisoning threat model; does it apply to misalignment? 🧵
all congressional speeches should be annotated with laugh tracks