一作ACL 2026Findings· Findings of ACL
EVADE : LLM -Based Explanation Generation and Validation for Error Detection in NLI
Longfei Zuo · Barbara Plank · Siyao Peng
摘要
High-quality datasets are critical for training and evaluating reliable NLP models. In tasks like natural language inference (NLI), human label variation (HLV) arises when multiple labels are valid for the same instance, making it difficult to separate annotation errors from plausible variation....