超越单轮：大型语言模型多轮交互的综述
Beyond Single-Turn: A Survey on Multi-Turn Interactions with Large Language Models
AI总结：
本文综述了大型语言模型在多轮交互中的评估与增强进展，探讨了多轮对话中上下文、连贯性、公平性和响应性等挑战，并总结了模型优化、外部整合和协作技术等方法。
摘要：
近年来，大型语言模型（LLMs）在单轮任务性能上取得了显著进展，但现实应用日益要求复杂的多轮交互。本文综述了评估和增强多轮LLM交互的最新进展。围绕以任务为导向的分类，涵盖数学、编程等领域的指令跟随，以及角色扮演、医疗、教育和对抗性jailbreak等场景的对话参与，系统地探讨了在长时间对话中维持上下文、连贯性、公平性和响应性的挑战。我们将现有基准和数据集组织成反映多轮对话评估演变景观的连贯类别，并回顾了广泛增强方法，包括模型中心策略（上下文学习、监督微调、强化学习和架构创新）、外部整合方法（记忆增强、基于检索的方法和知识图谱）以及协作交互的代理技术。最后，我们识别了未来研究的开放挑战和有前景的方向，以进一步提高多轮LLM交互的鲁棒性和有效性。
英文摘要：
Recent advances in large language models (LLMs) have substantially improved single-turn task performance, yet real-world applications increasingly demand sophisticated multi-turn interactions. This survey provides a comprehensive review of recent progress in evaluating and enhancing multi-turn LLM interactions. Centered on a task-oriented taxonomy-spanning instruction following in domains such as mathematics and coding, and conversational engagement in role-playing, healthcare, education, and adversarial jailbreak settings-we systematically examine the challenges of maintaining context, coherence, fairness, and responsiveness across prolonged dialogues. We organize existing benchmarks and datasets into coherent categories reflecting the evolving landscape of multi-turn dialogue evaluation, and review a broad spectrum of enhancement methodologies, including model-centric strategies (in-context learning, supervised fine-tuning, reinforcement learning, and architectural innovations), external integration approaches (memory augmentation, retrieval-based methods, and knowledge graphs), and agent-based techniques for collaborative interaction. Finally, we identify open challenges and promising directions for future research to further improve the robustness and effectiveness of multi-turn LLM interactions.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
