{"ID":2896728,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.07203","arxiv_id":"2507.07203","title":"State-Inference-Based Prompting for Natural Language Trading with Game NPCs","abstract":"Large Language Models enable dynamic game interactions but struggle with rule-governed trading systems. Current implementations suffer from rule violations, such as item hallucinations and calculation errors, that erode player trust. Here, State-Inference-Based Prompting (SIBP) enables reliable trading through autonomous dialogue state inference and context-specific rule adherence. The approach decomposes trading into six states within a unified prompt framework, implementing context-aware item referencing and placeholder-based price calculations. Evaluation across 100 trading dialogues demonstrates \u003e97% state compliance, \u003e95% referencing accuracy, and 99.7% calculation precision. SIBP maintains computational efficiency while outperforming baseline approaches, establishing a practical foundation for trustworthy NPC interactions in commercial games.","short_abstract":"Large Language Models enable dynamic game interactions but struggle with rule-governed trading systems. Current implementations suffer from rule violations, such as item hallucinations and calculation errors, that erode player trust. Here, State-Inference-Based Prompting (SIBP) enables reliable trading through autonomo...","url_abs":"https://arxiv.org/abs/2507.07203","url_pdf":"https://arxiv.org/pdf/2507.07203v1","authors":"[\"Minkyung Kim\",\"Junsik Kim\",\"Hwidong Bae\",\"Woongcheol Yang\",\"Sangdon Park\",\"Sohee Bae\"]","published":"2025-07-09T18:24:47Z","proceeding":"cs.AI","tasks":"[\"cs.AI\"]","methods":"[\"Language Model\"]","has_code":false}