Research · paper
Adaptive Prompting for Continual Relation Extraction
A paper-review template for tracking method, datasets, results, limitations, and personal takeaways.
Problem
Continual relation extraction asks a model to learn new relation types without forgetting previous ones.
Research question
Can adaptive prompts preserve task-specific knowledge while allowing the model to absorb new relation classes?
Method
Track the prompt design, replay strategy, classifier changes, and evaluation protocol.
Limitations
Check whether gains come from the prompting method itself, additional memory, larger backbones, or dataset-specific behavior.