Shankar Padmanabhan

@shankarpad8

Third year NLP PhD student at Cornell, fellow @DatologyAI, NSF GFRP fellow, he/him

New York City

shankarp8.github.io

Joined February 2023

165 Following

144 Followers

Replies Reposts Media

Pinned
Shankar Padmanabhan
@shankarpad8
Mar 23
1/5 How do we update a model trained in 2025 with new world knowledge from 2026? ⚠️Continued training will undo skills learned by LLMs during post-training, e.g. instruction-following/math/code. 🤝Our method DiSC updates LLMs with new knowledge while preserving existing skills!
1
Shankar Padmanabhan
@shankarpad8
Oct 14
Most of what the field knows about training data lives in ablations that never get published. Labs is our attempt to change that; excited to start sharing what we've been working on!
DatologyAI
@datologyai
Oct 14
Today we’re launching Datology Labs: a new commitment to open and collaborative research into the science of data. Research has been a core pillar of DatologyAI from the start. Now we’re partnering with researchers at universities and across industry to take it further.
Shankar Padmanabhan
@shankarpad8
Oct 9
Data curation is the most underrated lever in model training, excited to see it finally become accessible to everyone! Huge congrats to the team!
DatologyAI
@datologyai
Oct 9
Today, we’re excited to launch the Datology Curation Studio, customizable frontier data curation for every team training its own models. Data quality is the ultimate compute multiplier, and Curation Studio makes it easy for everyone.
Shankar Padmanabhan
@shankarpad8
Jul 7
I’ll be presenting our 🌟Spotlight🌟 paper on continual learning at ICML! 📌 Wednesday (July 8), 5:00–6:45 PM, Poster Session 5 Come say hi if you’re around, happy to talk continual learning or anything else!
Shankar Padmanabhan
@shankarpad8
Mar 23
1/5 How do we update a model trained in 2025 with new world knowledge from 2026? ⚠️Continued training will undo skills learned by LLMs during post-training, e.g. instruction-following/math/code. 🤝Our method DiSC updates LLMs with new knowledge while preserving existing skills!
Shankar Padmanabhan
@shankarpad8
May 20
This paper is now a spotlight at ICML 2026!
Shankar Padmanabhan
@shankarpad8
Mar 23
1/5 How do we update a model trained in 2025 with new world knowledge from 2026? ⚠️Continued training will undo skills learned by LLMs during post-training, e.g. instruction-following/math/code. 🤝Our method DiSC updates LLMs with new knowledge while preserving existing skills!
1

Shankar Padmanabhan

Shankar Padmanabhan’s posts