Brief

Researchers have developed a new agentic large language model (LLM) framework designed to improve the classification of Harmonized Tariff Schedule (HTS) codes, which are crucial for international trade and customs. The framework incorporates multi-agent retrieval, semantic search of tariff documents, and a consensus-based validation system with element-wise voting and confidence estimation. While experimental results on a dataset of 3,300 product records indicate that precise 10-digit classification remains challenging for LLMs, the proposed system offers a more interpretable and accountable approach for maritime logistics and smart-port operations. AI

A new research paper introduces ACTION-RATING, a method to integrate clarification-seeking directly into the action space of hierarchical language agents. This formulation allows agents to compete between acting and asking for help at each decision point, leading to observable help-seeking behaviors. The study observed a shift from mandatory to opportunistic clarification, significantly improving Information-Seeking Effectiveness. AI