A knowledge-guided pipeline uses DeepSeek V3 to generate and self-verify 50K Chinese legal reasoning examples, and fine-tuned Qwen-2.5 models (0.5B-3B) into LawGPT, which outperforms prior legal-specific LLMs on four LawBench-style tasks.
Learning to predict charges for criminal cases with legal basis
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
LawGPT: Knowledge-Guided Data Generation and Its Application to Legal LLM
A knowledge-guided pipeline uses DeepSeek V3 to generate and self-verify 50K Chinese legal reasoning examples, and fine-tuned Qwen-2.5 models (0.5B-3B) into LawGPT, which outperforms prior legal-specific LLMs on four LawBench-style tasks.