complexity-aware

CoAL-RAG introduces a complexity-aware retrieval-augmented generation method for legal consultation that dynamically selects retrieval strategies based on question complexity, achieving significant improvements on Chinese and English legal benchmarks.

DecoSearch introduces a training-free framework for Text-to-SQL that routes queries based on complexity, using schema selection, decomposition into DAGs, RAG, and plan-level repair, achieving 70.53% on BIRD and 88.31% on Spider with a DeepSeek backbone.