Tag
This paper investigates whether context-injected fine-tuning improves small language models' ability to use retrieved law for legal question answering in Bangladesh. Using Qwen3.5 at 0.8B, 2B, and 4B scales, they find fine-tuning helps at smaller scales but not at 4B, and reduces language drift.
Proposes Decompose-and-Refine (DaR), a framework for statute-grounded legal question answering that decomposes complex questions into atomic sub-questions and generates parametric queries for precise statutory retrieval, showing improvements on the KoBLEX benchmark.
This paper studies temporal failure modes in LLM-based statutory question answering, including post-cutoff staleness and recency bias. It introduces a benchmark of 312 expert-validated German statutory QA pairs and evaluates LLMs under various inference settings.