idan shenfeld

@IdanShenfeld

Research @thinkymachines | Prev @MIT @GoogleDeepMind

Cambridge, MA

idanshenfeld.com

Joined December 2022

198 Following

2,208 Followers

Replies Reposts Media

Pinned
idan shenfeld
@IdanShenfeld
Jan 29
People keep saying 2026 will be the year of continual learning. But there are still major technical challenges to making it a reality. Today we take the next step towards that goal — a new on-policy learning algorithm, suitable for continual learning! (1/n)
51
idan shenfeld
@IdanShenfeld
Oct 3
SFT without forgetting? Possible if you make your data more on policy! Really nice work from @aakara
Aayush Karan
@aakaran31
Oct 2
SFT is not dead! 🥳 We found a way to make SFT rival current prevailing posttraining methods, often generalizing better and forgetting less than RL and OPSD. 🤯 Following our prior work on reasoning with sampling, we now introduce sampling to the posttraining stack. 1/n
GIF
1
idan shenfeld
@IdanShenfeld
Oct 2
Interesting new approach to catastrophic forgetting
Assaf Ben Kish
@abk_tau
Oct 2
(1/7) Humans can acquire many capabilities without (catastrophically) forgetting previous ones. Why can't neural networks do the same? Local Support Learning (LSL) takes us a step closer towards this goal. In LLMs of up to 7B parameters, LSL augments gradient-based training to
idan shenfeld
@IdanShenfeld
Sep 24
Sometimes we are so focused on LLMs that we miss how AI has improved in other use cases
Victor Butoi
@ion_barrel
Sep 24
X-rays are everywhere in medicine, but extracting general, quantitative anatomical information from them remains remarkably difficult. Announcing FleXray: an open-source, open-weight model for zero-shot anatomical segmentation across the human body—from different regions,
GIF
1
idan shenfeld
@IdanShenfeld
Aug 19
Continual learning on your phone! super excited to see this adaptation of SDFT
Zhang-Wei Hong
@ZhangWeiHong9
Aug 18
Continual learning running on a phone. Not inference. Not a fine-tune you kick off with a dataset. A small language model (SLM) that keeps learning from its own experience while it's being used, with every gradient step on the handset. How it works (lin826.github.io/SLM-Online-SDF…):

idan shenfeld

idan shenfeld’s posts