LEGIT converts court judgments into hierarchical issue trees that act as expert rubrics to measure coverage and correctness of LLM-generated legal reasoning traces, showing that RAG and rubric-based RL provide complementary gains.
Refer to Figure 10 for the original version of the data and LLM responses
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Evaluating Legal Reasoning Traces with Legal Issue Tree Rubrics
LEGIT converts court judgments into hierarchical issue trees that act as expert rubrics to measure coverage and correctness of LLM-generated legal reasoning traces, showing that RAG and rubric-based RL provide complementary gains.