We are excited to have a poster presentation for our new work "BLISS: A Lightweight Bilevel Influence Scoring Method for Data Selection in Language Model Pretraining" in #ICML2026 at Hall A (#2607), COEX Convention & Exhibition Center in Seoul on July 8th (5:00-6:45 p.m. KST).
[2/2] We propose a new bilevel formulation with a smoothed Top-k regularizer for coreset selection in continual learning, and provide the theoretical guarantee and interesting experimental results ( https://t.co/alOgeeqynI). Looking forward to your attention and participation
[1/2] We are excited to have a poster presentation for our new work "Bilevel Coreset Selection in Continual Learning: A New Formulation and Algorithm" in #Neurips2023 at Great Hall B1+B2 (#1023), Ernest N. Morial Convention Center on Dec 12 (5:15-7:15 p.m. CST).