ERGO：多輪語言模型生成優化中的熵引導重置策略

摘要

大型語言模型（LLMs）在資訊逐步呈現的多輪對話中，性能會顯著下降。考慮到多輪對話是日常與LLMs互動的特徵，這種性能下降對實際應用構成了嚴峻挑戰。我們假設，模型不確定性的突然增加標誌著多輪LLM互動中的不對齊，並利用這一洞察力來動態重新對齊對話上下文。我們引入了ERGO（基於熵的生成優化重置），它通過對下一個詞元分佈的香農熵持續量化內部不確定性，並在檢測到熵的急劇上升時觸發自適應提示整合。通過將不確定性視為首要信號而非需要消除的干擾，ERGO接納了語言和建模中的變異性，並對不確定性進行表示和響應。在逐步揭示指令的多輪任務中，ERGO相比標準基線平均提升了56.6%的性能，提高了24.7%的適任能力（峰值性能能力），並減少了35.3%的不可靠性（性能變異性），這表明對不確定性有意識的干預能夠提升對話AI的準確性和可靠性。

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：多輪語言模型生成優化中的熵引導重置策略

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

摘要

Support