We present LLM-INSTRUCT, the winning system for the UZH Shared Task at ArgMining 2026, focusing on paragraph-level argument mining in UN and UNESCO resolutions. The task requires paragraph-type classification, prediction of a subset of 141 official tags, and directed relation prediction under a strict JSON schema setting using only open-weight models up to 8B parameters.
We frame the task as constrained structured prediction. The system first narrows the candidate tag space with metadata-aware dense retrieval, then applies constrained decoding with per-dimension caps, escalates only uncertain cases to a three-agent debate branch, and finally validates the output schema. On the official leaderboard, LLM-INSTRUCT ranked 1st overall, with 1st in F1 and 5th in LLM-as-a-Judge. During development, our configuration search further improved Task 1b Micro-F1 from 35.83% to 40.08% while keeping the internal Task 2 score at 4.421. The main lesson is simple: reducing the decision space before generation improves both accuracy and submission robustness.
Our code and supporting scripts are publicly available at: LLM-INSTRUCT GitHub.
Blogger's Review: LLM-INSTRUCT showcases the efficiency and accuracy of metadata-aware retrieval and constrained decoding techniques in paragraph-level argument mining, providing significant insights for related research. Its methodology is a valuable reference for other tasks, especially in dealing with complex structured data.