TDR: Task-Decoupled Retrieval with Fine-Grained LLM Feedback for In-Context Learning

Open in new window