This paper describes Team CV’s systems forSemEval-2026 Task 4: Narrative Story Sim-ilarity and Narrative Representation Learn-ing (Hatzel et al., 2026). For Track A (com-parative judgment), we explore five prompt-ing strategies—zero-shot, chain-of-thought,structured feature extraction, pairwise scor-ing, and few-shot—and QLoRA fine-tuningof smaller models. For Track B (narrativeembeddings), we benchmark twelve dedicatedtext embedding models of varying dimen-sionality (384–4096) spanning open-source(E5-Large-v2, BGE, GTE, Qwen3 Embed-ding) and closed-source (OpenAI, Gemini,Mistral) families, and fine-tune Qwen3 Em-bedding 4B on task-specific triples. Few-shot prompting with Qwen-2.5 7B (64.00%)outperforms all fine-tuned variants (best57.50%) on Track A; scaling to LLaMA-3.3-70B yields 75.00%. On Track B, Ope-nAI text-embedding-3-large (3072-d) achieves the best dev accuracy (67.00%),while fine-tuning Qwen3 Embedding 4B(2560-d) on synthetic triples slightly de-creases accuracy. Our final submission—LLaMA-3.3-70B (3-shot) for Track A andtext-embedding-3-large for Track B—achieves 70.75% and 64.50%, exceeding theGPT-4o-mini and STORY-EMB baselines respec-tively.

United Nations and UNESCO resolutions encode complex collective reasoning through highly structured preambles and operative clauses. Reconstructing this implicit argumentative structure is a challenging natural language processing task. This paper describes our submission to the UZH Shared Task at the ArgMining Workshop 2026. Adhering to the strict constraint of using open-weight models with ≤ 8B parameters, we propose a highly efficient, modular pipeline built entirely upon the Qwen-2.5-7B-Instruct architecture. To address Subtask 1, we decouple the problem: employing a 4-bit quantized LoRA adapter via the Unsloth framework for paragraph type classification, alongside a dimension-chunked zero-shot approach to assign multi-label tags from a complex 141-class taxonomy. For Subtask 2, we introduce a novel Type-Informed Chain-of-Thought (CoT) methodology that leverages predicted structural metadata as formal constraints to extract argumentative links. To overcome the inherent context window limitations of sub-8B models on extended documents, we implement a multi-pass context recovery pipeline. Our system successfully processes the entire test set of 2,959 paragraphs, ultimately securing an overall Final Rank of 8th in the shared task (ranking 6th on the F1 leaderboard and 8th on the LLM-Judge leaderboard).