Story
arxiv_cs_lg · Nov 9, 2026 · paper
Source brief
Anchor Behavior, Not Parameters: Rehearsal-Free Continual Learning with a Single Shared Prompt
arxiv.orgNov 9, 2026
original source linked
In brief
Prompt-based continual learning adapts a frozen pre-trained vision transformer by tuning a small set of prompt tokens. Recently, one-prompt methods which share one single prompt across all tasks make this paradigm max...
Feed lens
evaluation