Rohan Varma

@rvarm1

@reflection_ai | prev: Jump Trading, Meta Superintelligence, @PyTorch

New York, NY

rohanvarma.io

Joined June 2014

724 Following

1,312 Followers

Replies Reposts Media

Pinned
Rohan Varma
@rvarm1
Oct 5
Fun time cooking this up with an awesome team! More to come soon 🚀🚀🚀
Reflection
@reflection_ai
Oct 5
Introducing Beam: a highly efficient agentic open model with 501B total parameters and 23B active. - Frontier reasoning efficiency - Advances the Western open frontier on coding & agentic tasks - Trained end-to-end from scratch Full weights release this month. Learn more about
5
Rohan Varma
@rvarm1
Oct 6
Was amazing to work with @achowdhery and many others on this! A nice peek into our pretraining science + infra recipe:
Aakanksha Chowdhery
@achowdhery
Oct 6
Highlighting noteworthy details from the rigorous science of scaling that the pretraining team followed to build the strong reasoning base model for the Beam OSS release.
Quote
Reflection
@reflection_ai
Oct 5
Introducing Beam: a highly efficient agentic open model with 501B total parameters and 23B active. - Frontier reasoning efficiency - Advances the Western open frontier on coding & agentic tasks - Trained end-to-end from scratch Full weights release this month. Learn more about
Rohan Varma
@rvarm1
Jul 24
Helping make the AI ecosystem more open has been the lion’s share of my career, from PyTorch to Llama @AIatMeta to the open models we’re now building @reflection_ai. Excited for an open weight future 🚀
Jensen Huang
@JensenHuang
Jul 24
For my first post, I’m sharing a letter @nvidia signed on why open models matter. AI will transform every industry, power every company, and be built by every country. Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.
1
Rohan Varma
@rvarm1
Jul 1
The day when etched took over the TL
Rohan Varma
@rvarm1
May 22
We are in the good timeline
Dwarkesh Patel
@dwarkesh_sp
May 22
New blackboard lecture w @reinerpope How do chips actually work – starting with basic logic gates, and working up to why GPUs, TPUs, FPGAs, and the human brain each look the way they do. 0:00:00 – Building a multiply-accumulate from logic gates 0:16:20 – Muxes and the cost of
00:00