ERGO：多轮语言模型生成优化中的熵引导重置策略

摘要

大型语言模型（LLMs）在信息逐步呈现的多轮对话中表现显著下降。鉴于多轮对话是日常生活中与LLMs互动的典型特征，这种性能退化对实际应用构成了严峻挑战。我们假设，模型不确定性的急剧增加标志着多轮LLM交互中的错位，并利用这一洞察力动态调整对话上下文。我们引入了ERGO（基于熵的生成优化重置机制），它通过香农熵持续量化下一令牌分布的内部不确定性，并在检测到熵值急剧上升时触发自适应提示整合。ERGO将不确定性视为首要信号而非需消除的干扰，接纳语言和建模中的变异性，代表并响应不确定性。在逐步揭示指令的多轮任务中，ERGO相较于标准基线实现了56.6%的平均性能提升，提升了24.7%的峰值性能能力，并减少了35.3%的性能波动，证明了不确定性感知干预能够提升对话AI的准确性和可靠性。

English

Large Language Models (LLMs) suffer significant performance degradation in multi-turn conversations when information is presented incrementally. Given that multi-turn conversations characterize everyday interactions with LLMs, this degradation poses a severe challenge to real world usability. We hypothesize that abrupt increases in model uncertainty signal misalignment in multi-turn LLM interactions, and we exploit this insight to dynamically realign conversational context. We introduce ERGO (Entropy-guided Resetting for Generation Optimization), which continuously quantifies internal uncertainty via Shannon entropy over next token distributions and triggers adaptive prompt consolidation when a sharp spike in entropy is detected. By treating uncertainty as a first class signal rather than a nuisance to eliminate, ERGO embraces variability in language and modeling, representing and responding to uncertainty. In multi-turn tasks with incrementally revealed instructions, ERGO yields a 56.6% average performance gain over standard baselines, increases aptitude (peak performance capability) by 24.7%, and decreases unreliability (variability in performance) by 35.3%, demonstrating that uncertainty aware interventions can improve both accuracy and reliability in conversational AI.

ERGO：多轮语言模型生成优化中的熵引导重置策略

ERGO: Entropy-guided Resetting for Generation Optimization in Multi-turn Language Models

摘要

Support