Ashley Edwards
@ashrewards
Followers
1,038
Following
241
Media
1
Statuses
72
Research @runwayml . Past: Google DeepMind, Uber AI Labs, Georgia Tech
London, England
Joined February 2024
Don't wanna be here?
Send us removal request.
Explore trending content on Musk Viewer
#KawalPutusanMK
• 2764619 Tweets
#TolakPolitikDinasti
• 1619535 Tweets
#TolakPilkadaAkal2an
• 1521779 Tweets
#BeyondTheVinesXFreenBecky
• 357294 Tweets
FB JOIN OPENING BTV HOUSE
• 356094 Tweets
Nobara
• 254045 Tweets
BATALKAN BUKAN TUNDA
• 234477 Tweets
Mulyono
• 191603 Tweets
JANGAN LENGAH
• 152419 Tweets
Nigerians
• 119561 Tweets
Bolt
• 78775 Tweets
South Africans
• 75676 Tweets
BOTO PARA SA SB19
• 61567 Tweets
鯉登少尉
• 42664 Tweets
ポケ森サ終
• 37231 Tweets
チャイルドシート
• 35334 Tweets
首都高バトル
• 26127 Tweets
BTV X FORCEBOOK
• 25016 Tweets
買い切り
• 23169 Tweets
SF x PONDPHUWIN
• 22003 Tweets
中川大志
• 21706 Tweets
本人確認
• 17846 Tweets
南京大虐殺
• 15426 Tweets
GCSE
• 14934 Tweets
This was such a fun and rewarding project to work on. Amazing job by the team! The most exciting thing for me is that we were able to achieve this without using a single doggone action label, which believe me, was not easy!
I am really excited to reveal what
@GoogleDeepMind
's Open Endedness Team has been up to 🚀. We introduce Genie 🧞, a foundation world model trained exclusively from Internet videos that can generate an endless variety of action-controllable 2D worlds given image prompts.
145
572
3K
10
17
114
Also, I created a Twitter account after 3-4 years just to share how much I’ve loved working on this project with a really amazing team 🫶
3
0
24
Thanks for having me! It was fun to chat :D
Today, we're joined by
@ashrewards
, a member of technical staff at
@runwayml
, to discuss Genie: Generative Interactive Environment. We explore the motivations behind Genie, the challenges in data acquisition for RL, the model architecture, training strategies, benchmarks,
1
8
33
0
1
22
I’ll be hanging out in SF and NY the next couple of weeks though sticking around in London (the weather hasn’t gotten to me yet)
1
0
22
Thank you
@nathanbenaich
! Really looking forward to it :D
Happy to host
@ashrewards
of
@GoogleDeepMind
at
@raais
2024!
Ashley works on reinforcement learning and foundational world models. She was part of the team that created Gato, a multi-modal, multi-task, multi-embodiment agent that was able to perform a diverse range of tasks,
2
6
32
1
1
19
sammee!! 🫶
Working on Genie with
@GoogleDeepMind
's Open-Endedness Team and our colleagues has been a highlight of my career. Thank you Jake Bruce,
@MichaelD1729
,
@ashrewards
,
@jparkerholder
,
@YugeTen
,
@edwardfhughes
, Matthew Lai,
@aditimavalankar
,
@richie_internet
, and many others.
1
4
82
0
0
13
Look at Doris go! :D
Getting our pets to move around was definitely one of the highlights of the project
It has been a dream to work on Genie with such fantastic people, I’ve learned so much from all of them. We've also had a lot of fun, for example, using our model trained on platformers to convert random pictures of our pets into playable worlds 🤯🐶
0
2
16
1
0
9
Thanks Nando! 🫶
0
0
6
Can’t wait!! 🧞♂️
1
0
6
Can’t wait!! 🎉
We are hosting our first Research and Art RNA session in London! 🇬🇧
If you want to apply to speak or know someone who would be great to have, let us know.
12
7
96
0
0
6
@nathanbenaich
I'll mainly be discussing our recent work Genie and how it enables us to learn playable environments from videos alone 🧞
🧞 Genie: Generative Interactive Environments
A Foundation Model for Playable Worlds
sites.google.com
0
0
5
@jparkerholder
I’m new here so not sure of the limit of twitter sap is — but ditto mate couldn’t have done this without you! (Re: section 6.4 )
0
0
5
@jeffclune
@jparkerholder
@akbirkhan
@MichaelD1729
@_rockt
Getting the characters in these sketches to move around was probably one of my favorite parts of the whole project :)
0
1
4
@NandoDF
@GoogleDeepMind
Thank you Nando! It truly is an amazing team, and it’s been great to have your support throughout the project :D
1
0
4
1
0
3
1
0
3
This workshop's going to be so fun -- can't wait!
We are pleased to announce the first *controllable video generation* workshop at
@icmlconf
2024! 📽️📽️📽️
We welcome submissions that explore video generation via different modes of control (e.g. text, pose, action).
Deadline: 31st May AOE
Website:
1
9
47
1
0
2
@jparkerholder
haha thanks for the shout out :D. Also reminds me a bit of this work from
@scottniekum
's group from back in the day . Cool to see this line of work picking back up!
0
0
2
@YugeTen
Hahaha I’m just stubborn but it took you and everyone in the team to get it to actually work! And I agree it’s been so fun working together 🫶
0
0
2