Archaeology at WE-2026 PARSEME 2.0 Subtask 1 and 2: Parsing is for Encoders, Paraphrasing is for LLMs

AbstractThis paper presents our approach to the PARSEME 2.0 Shared Task on Romanian, covering both Identification (Subtask 1) and Paraphrasing (Subtask 2). While Large Language Models (LLMs) excel at semantic generation, we hypothesize that they lack the structural precision required for MWE identification, leading to "boundary hallucinations" that compromise downstream simplification. Our Rank 1 results on Romanian confirm this: a specialized encoder (RoBERT) using standard sequence labeling outperforms both few-shot LLMs and complex structural parsers (MTLB-STRUCT). This justifies our proposed pipeline: using encoders as precise “pointers” to guide the generative power of LLMs.