AI Navigate
インサイト最新記事AI大全カオスマップ

METRO: Towards Strategy Induction from Expert Dialogue Transcripts for Non-collaborative Dialogues

arXiv cs.CL / 4/14/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research
Read original →
共有:

Key Points

  • The paper proposes METRO, a method that uses large language models to automatically induce both dialogue strategy actions and planning logic from expert non-collaborative dialogue transcripts, reducing reliance on manual codification.
  • METRO represents expert knowledge as a hierarchical “Strategy Forest,” combining short-term response patterns with longer-horizon strategic foresight via branching structure.
  • Experiments on two benchmarks indicate METRO achieves average performance gains of about 9%-10% over existing approaches.
  • The authors find that METRO’s effectiveness is linked to strategic behavioral diversity and foresight, and that the learned strategies transfer robustly across different tasks.
  • The authors provide an open-source codebase at the linked GitHub repository to support reproducibility and further development.

Abstract

Developing non-collaborative dialogue agents traditionally requires the manual, unscalable codification of expert strategies. We propose \ours, a method that leverages large language models to autonomously induce both strategy actions and planning logic directly from raw transcripts. METRO formalizes expert knowledge into a Strategy Forest, a hierarchical structure that captures both short-term responses (nodes) and long-term strategic foresight (branches). Experimental results across two benchmarks show that METRO demonstrates promising performance, outperforming existing methods by an average of 9%-10%. Our further analysis not only reveals the success behind METRO (strategic behavioral diversity and foresight), but also demonstrates its robust cross-task transferability. This offers new insights into building non-collaborative agents in a cost-effective and scalable way. Our code is available at https://github.com/Humphrey-0125/METRO.

関連おすすめサービス

※当サイトはアフィリエイト広告を利用しています
🎧

Notta搭載AI議事録イヤホン ZENCHORD1

AI時代の仕事術。Notta搭載で会議の議事録を自動生成するスマートイヤホン。

🎙️

AI搭載ボイスレコーダー Plaud

世界100万人が愛用。AIで文字起こし・要約を自動化するボイスレコーダー。

🖼️

画像高画質化AIツール Aiarty Image Enhancer

AIで画像を高画質化。写真・イラストを簡単にアップスケール。