PEARL uses an input-agnostic global prompt with a negative-feedback adaptive momentum update to reduce catastrophic forgetting in pre-trained-model class-incremental learning, achieving state-of-the-art average accuracy of 86.41% across six benchmarks.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
PEARL: Input-Agnostic Prompt Enhancement with Negative Feedback Regulation for Class-Incremental Learning
PEARL uses an input-agnostic global prompt with a negative-feedback adaptive momentum update to reduce catastrophic forgetting in pre-trained-model class-incremental learning, achieving state-of-the-art average accuracy of 86.41% across six benchmarks.