Natively unlearnable language models highlighted ahead of COLM 2026 workshop talk
A post invites readers to Gaurav's oral presentation at the HAIPS workshop, pointing to new architectures and unlearning.
TLDR
A user invites people interested in new architectures and effective unlearning to Gaurav's oral presentation at the HAIPS workshop at COLM 2026. The user calls natively unlearnable language models “a pretty cool idea.”
Combined views
—
1 Source, first seen ago
