tensor_rotator's profile picture. Inference @AnthropicAI, prev Gemini @Google, prev prev PhD @UTAustin

Alek Dimitriev

@tensor_rotator

Inference @AnthropicAI, prev Gemini @Google, prev prev PhD @UTAustin

ปักหมุด

Imagine being called 'former salesforce exec' instead of 'ML researcher with 170,000+ citations'

tensor_rotator's tweet image. Imagine being called 'former salesforce exec' instead of 'ML researcher with 170,000+ citations'

Alek Dimitriev รีโพสต์แล้ว

Turns out our version of the Romans eating lead isn't microplastics, it's also just eating lead.

New potential explanation for “what’s happening in America in 2025”

robinsonmeyer's tweet image. New potential explanation for “what’s happening in America in 2025”


Alek Dimitriev รีโพสต์แล้ว

they can't keep getting away with this

AndyAyrey's tweet image. they can't keep getting away with this

Haiku 4.5 is out and it shatters the SWE bench to cost Pareto frontier, please try it!

Introducing Claude Haiku 4.5: our latest small model. Five months ago, Claude Sonnet 4 was state-of-the-art. Today, Haiku 4.5 matches its coding performance at one-third the cost and more than twice the speed.

claudeai's tweet image. Introducing Claude Haiku 4.5: our latest small model.

Five months ago, Claude Sonnet 4 was state-of-the-art. Today, Haiku 4.5 matches its coding performance at one-third the cost and more than twice the speed.


Alek Dimitriev รีโพสต์แล้ว

1/8 Second Order Optimizers like SOAP and Muon have shown impressive performance on LLM optimization. But are we fully utilizing the potential of second order information? New work: we show that a full second order optimizer is much better than existing optimizers in terms of…

ShamKakade6's tweet image. 1/8 Second Order Optimizers like SOAP and Muon have shown impressive performance on LLM optimization. But are we fully utilizing the potential of second order information? New work: we show that a full second order optimizer is much better than existing optimizers in terms of…

Alek Dimitriev รีโพสต์แล้ว
dwarkesh_sp's tweet image.

16th in nation

dwarkesh_sp's tweet image. 16th in nation


Alek Dimitriev รีโพสต์แล้ว

Porsche Macan is LA’s Honda Accord


This can be entirely explained by easier or harder PRs being given to Claude Code vs Codex?

Developers approved 74.3% of code written by Codex compared to 73.7% for Claude Code, according to data from more than 300,000 pull requests collected by startup Modu Sourcegraph's Amp agent had the highest acceptance rate at 76.8%



Alek Dimitriev รีโพสต์แล้ว

Today we are launching InferenceMAX! We have support from Nvidia, AMD, OpenAI, Microsoft, Pytorch, SGLang, vLLM, Oracle, CoreWeave, TogetherAI, Nebius, Crusoe, HPE, SuperMicro, Dell It runs every day on the latest software (vLLM, SGLang, etc) across hundreds of GPUs, $10Ms of…

Going to be dropping something huge in 24 hours I think it'll reshape how everyone thinks about chips, inference, and infrastructure It's directly supported by NVIDIA, AMD, Microsoft, OpenAI, Together AI, CoreWeave, Nebius, PyTorch Foundation, Supermicro, Crusoe, HPE, Tensorwave,…



Time for a friends section in my library

tensor_rotator's tweet image. Time for a friends section in my library

Alek Dimitriev รีโพสต์แล้ว

The Scaling Era is out today. I'm actually surprised with how well this format works. Even better than my expectations. It's so interesting to read side-by-side how hyperscalar CEOs, AI researchers, and economists will answer the same question. Thank you to the @stripepress

I'm so pleased to present a new book with @stripepress: "The Scaling Era: An Oral History of AI, 2019-2025." Over the last few years, I interviewed the key people thinking about AI: scientists, CEOs, economists, philosophers. This book curates and organizes the highlights across…

dwarkesh_sp's tweet image. I'm so pleased to present a new book with @stripepress: "The Scaling Era: An Oral History of AI, 2019-2025."

Over the last few years, I interviewed the key people thinking about AI: scientists, CEOs, economists, philosophers. This book curates and organizes the highlights across…


Alek Dimitriev รีโพสต์แล้ว

Ever wondered what CAN'T be transformed by Transformers? 🪨 I wrote a fun blog post on finding "fixed points" of your LLMs. If you prompt it with a fixed point token, the LLM is gonna decode it repeatedly forever, guaranteed. There's some connection with LLMs' repetition issue.

liujc1998's tweet image. Ever wondered what CAN'T be transformed by Transformers? 🪨

I wrote a fun blog post on finding "fixed points" of your LLMs. If you prompt it with a fixed point token, the LLM is gonna decode it repeatedly forever, guaranteed.

There's some connection with LLMs' repetition issue.

Alek Dimitriev รีโพสต์แล้ว

What is intelligence? What will it take to create AGI? What happens once we succeed? The Scaling Era: An Oral History of AI, 2019–2025 by @dwarkesh_sp and @g_leech_ explores the questions animating those at the frontier of AI research. It’s out today: press.stripe.com/scaling


Alek Dimitriev รีโพสต์แล้ว
damintoell's tweet image.

NEW - Vyacheslav Leontyev, the 87-year-old head of the Pravda publishing house, died after falling 70 feet from his Moscow apartment window — Daily Mail

disclosetv's tweet image. NEW - Vyacheslav Leontyev, the 87-year-old head of the Pravda publishing house, died after falling 70 feet from his Moscow apartment window — Daily Mail


Alek Dimitriev รีโพสต์แล้ว

The SMS number-warming spam has gotten out of control.


Alek Dimitriev รีโพสต์แล้ว

the lion doesn’t concern himself with numeric stability


Alek Dimitriev รีโพสต์แล้ว

ughhhhh i didn’t fix my entire life this weekend FUCK


Alek Dimitriev รีโพสต์แล้ว

It’s crazy how far Apple has fallen

ZenFuturist's tweet image. It’s crazy how far Apple has fallen
ZenFuturist's tweet image. It’s crazy how far Apple has fallen

Loading...

Something went wrong.


Something went wrong.