Surya Bhupatiraju @suryabhupa Twitter profile

Pinned Tweet

Surya Bhupatiraju

2 months

I am absolutely thrilled to announce the release of Gemma 2! Today, we're releasing both pre-trained-only and fully post-trained 9B and 27B models. The full technical report is here: and it's live *right now* on .

Google AI Studio | Gemini API | Google for Developers | Google AI for Developers

Get started with the Gemini API on Google AI Studio. Quickly develop prompts for Gemini 1.5 Flash and 1.5 Pro with 2 million token context window.

ai.google.dev

21

50

234

Last Seen Profiles

@bieuaguilar

@yeongyuuie

@tujiangbaby

@GEN22OnSloane

@bokeplokalmalam

@PP_JavierArenas

@zalimsingh_0987

@lightingmccussy

@raysucassio

@umummmx01

@penyukastw21

@lyorteg

@stw_pdg

@TupSoftball

@submissivehunt

@mizukidayo_0408

@vong_mala25111

@lifewithjaidaa

@AmatyaPhd

@SolKongz

@Comptinator

@NextBanger

@squriby01

@OrgIAC

@yaho_ns

@NrseHuckleberry

@vivivivit_cs

@YabeeBear

@tugoverta

@BethUnderwood9

@FelipheDia23438

@mimayfair

@jandakembangstw

@bokeplokalmalam

@ayieko_onyango

@kaysinmusic

Surya Bhupatiraju

@suryabhupa

6 months

Thrilled to see Gemma released today, loved working on post-training with the team!

Jeff Dean (@🏡)

@JeffDean

6 months

Introducing Gemma - a family of lightweight, state-of-the-art open models for their class, built from the same research & technology used to create the Gemini models. Blog post: Tech report: This thread explores some of the

106

824

4K

8

54

Surya Bhupatiraju

@suryabhupa

5 years

Thanks to all who came to our workshop in Exploration in RL! The videos, slides, and papers are now available: . Thanks again to our speakers and panelists @pabbeel Doina Precup, @white_martha , Emo Todorov, @RaiaHadsell , @pulkitology , and @jeffclune ! :)

Home

Thanks for attending! ERL 2019 has concluded. The videos are available here: Introduction, Keynote by Doina Precup, and Author Spotlights Invited Talk by Emo Todorov, Best Paper Awards, and Invited...

sites.google.com

1

11

56

Surya Bhupatiraju

@suryabhupa

6 years

Check some new work I got to work on with @catherineols and others!

Catherine Olsson

@catherineols

6 years

Our paper "Skill Rating for Generative Models" is now up! tl;dr: A new idea & proof-of-concept for evaluating generative models. Train a bunch of GANs. Have the generators "play against" all the discriminator snapshots. Rate them like chess players. 1/n

3

86

332

0

22

Surya Bhupatiraju

@suryabhupa

4 months

Gemma v1.1 Instruct 2B and “7B” :) are out! See @robdadashi ’s thread for details, featuring improvements in multi-turn, fixing some overly-chatty features, and new RL. Even more to come soon!

Robert Dadashi

@robdadashi

4 months

I am very happy to announce that Gemma 1.1 Instruct 2B and “7B” are out! Here are a few details about the new models: 1/11

13

70

375

1

3

21

Surya Bhupatiraju

@suryabhupa

6 years

Check out our newest paper! Perhaps biased policy gradients really are the future of RL...

George Tucker

@georgejtucker

6 years

We looked at the sources of variance in policy gradient estimators for some common continuous control tasks, and I was surprised by the results: .

2

39

108

0

3

16

Surya Bhupatiraju

@suryabhupa

6 years

We've launched our first DFL Fellowship to help support more people in creating high-quality ML curricula -- please apply and share! :)

Depth First Learning

@DepthFirstLearn

6 years

We’re thrilled to announce the DFL fellowship, generously funded by @JaneStreetGroup . Have curriculum ideas? We are offering 4 fellows a $4000 grant each to build a 6 week curriculum and run weekly on-line discussions. Learn more and apply at ! (1/3)

1

23

75

0

5

14

Surya Bhupatiraju

@suryabhupa

6 years

Ben Eysenbach and I organized a workshop at ICML 2018 and decided to write about our experience, what we learned, and what we would've tried differently -- check it out! We hope it helps anyone looking to try organizing a workshop: (also go vote!)

Hitchhiker’s Guide to Organizing an Academic Workshop

Introduction

medium.com

0

13

Surya Bhupatiraju

@suryabhupa

6 years

Check out our guide about TRPO that I helped co-write for @DepthFirstLearn ! Feedback and suggestions are welcome :)

Depth First Learning

@DepthFirstLearn

6 years

We just released our newest study guide! Learn all about TRPO from professors @kumarkagrawal and @suryabhupa → .

2

52

160

0

1

11

Surya Bhupatiraju

@suryabhupa

7 years

Just watched an AI named Ousia absolutely CRUSH a team of very qualified Quiz Bowl veterans at Quiz Bowl in the Human-Computer Question Answering competition track! #NIPS2017

0

4

12

Surya Bhupatiraju

@suryabhupa

7 years

@hardmaru Related: how different probability metrics are related cf "On Choosing and Bounding Probability Metrics" (2002):

0

11

Surya Bhupatiraju

@suryabhupa

6 years

@anishathalye presenting adversarial turtles at ICML 2018 with @logan_engstrom @andrew_ilyas @antimatter15 !

0

2

10

Surya Bhupatiraju

@suryabhupa

6 years

Our workshop, Exploration in RL, is starting momentarily in Room T1 at #ICML — come by and hear some amazing speakers talk about solving exploration!

0

9

Surya Bhupatiraju

@suryabhupa

7 years

The new Google AI Residency has been announced! It's been simply fantastic so far -- please consider applying!

0

8

Surya Bhupatiraju

@suryabhupa

5 years

Huge thanks to the other co-organizers, including Ben Eysenbach, @shaneguML , @HarriLEdwards , @white_martha , @pyoudeyer , @EmmaBrunskill , @kenneth0stanley and sponsors @DeepMindAI and @GoogleAI !

0

2

6

Surya Bhupatiraju

@suryabhupa

8 months

@johnma2006 @_albertgu @tri_dao Super clean implementation!

0

4

Surya Bhupatiraju

@suryabhupa

6 years

Check out our new educational effort!

Depth First Learning

@DepthFirstLearn

6 years

Announcing ! We are building a repository of study guides targeting consequential papers. Check it out, learn something in-depth, and help us build the next one. @avitaloliver @suryabhupa @kumarkagrawal @cinjoncin

6

157

403

0

5

Surya Bhupatiraju

@suryabhupa

4 months

Gemma 1.1 improves on lmsys!

lmsys.org

@lmsysorg

4 months

Exciting news - the latest Arena result are out! @cohere 's Command R+ has climbed to the 6th spot, matching GPT-4-0314 level by 13K+ human votes! It's undoubtedly the **best** open model on the leaderboard now🔥 Big congrats to @cohere 's incredible work & valuable contribution

44

308

1K

0

6

Surya Bhupatiraju

@suryabhupa

6 months

@danielhanchen this is really thoughtfully done, thanks for surfacing all of these so clearly! we're on it :)

0

4

Surya Bhupatiraju

@suryabhupa

7 years

I recently gave a deep learning talk for prefrosh at MIT and I spent way too long on this slide (the rest was serious, I promise):

0

1

3

Surya Bhupatiraju

@suryabhupa

5 months

this is remarkable stuff, i can't imagine how much progress this team will continue to make!

Cognition

@cognition_labs

5 months

Today we're excited to introduce Devin, the first AI software engineer. Devin is the new state-of-the-art on the SWE-Bench coding benchmark, has successfully passed practical engineering interviews from leading AI companies, and has even completed real jobs on Upwork. Devin is

5K

11K

45K

0

4

Surya Bhupatiraju

@suryabhupa

7 years

A plethora of wonderful ideas/tricks/resources related to self-learning: .

0

4

Surya Bhupatiraju

@suryabhupa

6 years

Come see our poster at ICLR 2018! :)

George Tucker

@georgejtucker

6 years

We decompose the variance of a policy gradient estimator with an action dependent baseline which provides insights into previous methods and new opportunities for improvements. Workshop poster #2 at 11am today. @suryabhupa @shanegu @svlevine #ICLR2018

0

1

17

0

4

Surya Bhupatiraju

@suryabhupa

5 years

@SmithaMilli @catherineols @danieldewey @open_phil Huge congrats Smitha! :) Super well-deserved!!

0

3

Surya Bhupatiraju

@suryabhupa

7 years

@lishali88 @AmplifyPartners @pabbeel Congrats! Really exciting stuff :)

0

2

Surya Bhupatiraju

@suryabhupa

7 years

Such a wonderful read rec'd by @josephwandile : by @paulg -- would've been lovely to have read this years ago.

0

2

Surya Bhupatiraju

@suryabhupa

7 years

Press release of some of the work in neural program synthesis I did with researchers at @MSFTResearch : .

Deep Learning for Program Synthesis - Microsoft Research

By Rishabh Singh, Jacob Devlin, Abdelrahman Mohamed, and Pushmeet Kohli, Microsoft Research Despite the many advances in computing over the past decades, the actual process of writing computer...

www.microsoft.com

0

2

Surya Bhupatiraju

@suryabhupa

7 years

#Caffe2 's finally been released! My comments and TODOs from my last summer on AML are still there haha. Wonderful job to Yangqing et. al.!

0

1

Surya Bhupatiraju

@suryabhupa

7 years