Back to feed
arXiv cs.CL
arXiv cs.CL
7/24/2026
LLM-INSTRUCT at UZH Shared Task 2026: Constraint-Aware Retrieval and Selective Debate for Paragraph-Level Argument Mining

LLM-INSTRUCT at UZH Shared Task 2026: Constraint-Aware Retrieval and Selective Debate for Paragraph-Level Argument Mining

Short summary

LLM-INSTRUCT won the UZH Shared Task at ArgMining 2026 for paragraph-level argument mining in UN/UNESCO resolutions using open-weight models up to 8B parameters. The system narrows candidate tags via dense retrieval, applies constrained decoding, escalates uncertain cases to three-agent debate, and validates JSON schema output. Key lesson: reducing the decision space before generation improves both accuracy and robustness.

  • •Winning system for UZH ArgMining 2026 shared task on paragraph-level argument mining
  • •Uses dense retrieval, constrained decoding, and selective three-agent debate with open-weight models ≤8B
  • •Reducing decision space before generation is the core insight; code is open-sourced

Generated with AI, which can make mistakes.

#research-breakthrough#ai-agents#open-source
Read full article at arXiv cs.CL

Is this a good recommendation for you?

Comments

Failed to load comments. Please try again.

Explore more

NAVER LABS System Re-implementation for the IWSLT 2026 Instruction-Following Task
arXiv cs.CL
Transforming LLMs into Efficient Cross-Encoders via Knowledge Distillation for RAG Reranking
arXiv cs.CL
RAGthoven at SemEval-2026: Multi-Stage Humor Generation Pipeline Ties Frontier Model Baseline
arXiv cs.CL
L-MAD: A Systematic Evaluation of Multi-Agent Debate Structures in Legal Reasoning
arXiv CS.AI