Tag
The article introduces LearnActCoder, a role-aware error memory framework that adapts clinical coding agents by converting errors into structured lessons, improving CPT coding performance without model weight updates.
This paper investigates whether compact, task-specific bi-encoders fine-tuned on synthetic data from large language models can outperform general-purpose embeddings for clinical code retrieval in non-English languages, achieving state-of-the-art results on Spanish benchmarks CodiESP and DISTEMIST.