
Shashwat Singh
@shashwat_s19
AI researcher. Interested in the science of pretraining and interpretability.
ID: 1582018723799367680
17-10-2022 14:40:41
17 Tweet
49 Followers
85 Following

Happy to announce our latest research “Representation Surgery: Theory and Practice of Affine Steering", accepted at ICML! A joint work with roeeaharoni Jonathan Herzig @ryandcotterell Ponnurangam Kumaraguru “PK” and Shashwat Singh, done during an internship at Google AI.(1/8) arxiv.org/pdf/2402.09631



Did you think retraining without the deletion set is always enough (even if inefficient) for unlearning? Our paper Corrective Machine Unlearning showed a practical scenario where it is NOT, and is now accepted at TMLR! Work with Ameya P. @ ICML 2025, Amartya Sanyal, Oxford Torr Vision Group, Ponnurangam Kumaraguru “PK” 🧵👇


Our 📜Emergence of text semantics in CLIP Image encoder accepted UniReps Workshop NeurIPS Conference! W/ Sreeram Vennam* Shashwat Singh* Ani 📜: precog.iiit.ac.in/pubs/unireps-c… 🧵 🙏 Clémentine Dominé 🍊 Donato Crisostomi @ ICML Marco Fumero Zorah Lähner Luca Moschella Kim Stachenfeld (neurokim.bsky.social) #ProfGiri


