Share this page!

Enter URL or ID to Unroll

You can paste full URL like: https://x.com/threadreaderapp/status/1644127596119195649
or just the ID like: 1644127596119195649

How to get URL link on X (Twitter) App

  1. On the Twitter thread, click on or icon on the bottom
  2. Click again on or Share Via icon
  3. Click on Copy Link to Tweet
  4. Paste it above and click "Unroll Thread"!
  5. More info at Twitter Help
Rosanne Liu Profile picture
Rosanne Liu
@savvyRL
Mom. Cofounded & running @ml_collective. Host of Deep Learning Classics & Trends. Research at Google DeepMind. DEI/DIA Chair of ICLR & NeurIPS.
Oct 9, 2024 • 10 tweets • 3 min read
Last day of the very special @COLM_conf !! Surprise, surprise, I am actually here to present a poster, than just tweet 😆

Stop by poster #3 this afternoon if you want to learn about training LMs entirely, and from scratch, on knowledge graphs! Why, How and What we learned. Image @COLM_conf First, why? As old-fashioned ML researchers we were just very frustrated how **it's never clear what knowledge content is consumed in LM training**. The training data of regular LMs are huge, messy, inaccessible, ambiguous, & do not have a clear boundary of knowledge separation.
Jan 24, 2023 • 7 tweets • 6 min read
Now that we can write Tiny Papers @iclr_conf, what should we write about?

I'd like to invite all established researchers to contribute Tiny Ideas as inspirations, seeds for discussions & future collaborations! #TinyIdeasForTinyPapers

I'll start. Note: bad ideas == good starts. 1. Calibrate-before-train: before training every model with *data*, train them with noise to calibrate: loss function is to make sure they output "chance probability" — calibrate a model to be as neutral as possible before training starts. Does it help? Why or why not?
Jun 25, 2022 • 6 tweets • 3 min read
A quick thread on "How DALL-E 2, Imagen and Parti Architectures Differ" with breakdown into comparable modules, annotated with size 🧵
#dalle2 #imagen #parti

* figures taken from corresponding papers with slight modification
* parts used for training only are greyed out A compilation of model architecture diagrams of three recent By now we know that
- DALL-E & Imagen = diffusion; Parti = autoregressive
- Imagen & Parti use generic text encoders; DALLE uses CLIP enc

But in fact, one version of Imagen also used CLIP, one version of DALL-E also had AR prior. So there are more connections than it seemed.
Dec 13, 2020 • 20 tweets • 10 min read
Favorite #NeurIPS2020 presentations and posters this year

PS: heavily biased by what I happened to catch and whom I happened to talk to
PPS: still catching up on talks so the list is rather incomplete and I'd hope to grow
PPPS: with contributions from @ml_collective members [Talk] No. 1 has to go to -- keynote talk by @isbellHFh @mlittmancs et al simply brilliant 🎉🎉
slideslive.com/38935825/you-c…