An implementation of paper "Expectation-Guided Reasoning Verification: A Novel Cognition-Empowered Self-Reasoning Framework for Reasoning Refinement of Large Reasoning Models"
- Our main models (Expectation-Guided Knowledge Graph Generaion & Reasoning Knowledge Graph Verification) are in "main_model"
- Used datasets are in "data"
- Data processing codes (entity construction, entity description acquisition, and embedding) are in "data preprocessing"
- Evaluation metrics (BERTScore and LLM-Judge) are in "evaluation"
- All baselines are in "baselines"