arXiv cs.AI· Hannah Li·· 1 天前AI 评分31
当双方都在适应：评估交互式 LLM 系统的挑战
When Both Sides Adapt: The Challenges of Evaluating Interactive LLM Systems
AI 导读
研究指出，交互式 LLM 系统中用户会根据模型回复调整提示词策略与任务难度，使模型回复反过来影响后续评估所用的提示词集合。作者区分了共同分布效应与诱导分布效应两类估计量，证明离线基准评测与线上 A/B 测试分别收敛于这两类效应，可能导致模型排名反转与部署建议分歧，并给出了排名保证一致的充分条件。用户适应性还会增大估计方差，并构成 reward hacking、新奇效应等常见评估难题的共同机制。
正文 · 原文
Abstract:In interactive systems where users engage with LLMs, users are adaptive and, after observing LLM responses, can change their prompting strategy, as well as the type and difficulty of subsequent tasks. Thus model responses can influence the set of prompts on which the model is subsequently evaluated. We study the implications of this user adaptivity and prompt endogeneity for the evaluation of LLMs. We define two classes of estimands for comparing models: the common-distribution effect, which compares model performance on the same distribution of prompts, and the induced-distribution effect, which compares each model on the prompt distribution generated through user interaction. We show that offline benchmark evaluation and live A/B testing target different quantities, with the standard difference-in-means estimators for the two methods converging to the common-distribution effect and the induced-distribution effect, respectively. This distinction can affect decision-making: the two comparisons can rank models differently and lead to discrepancies in deployment recommendations. We characterize when such ranking reversals occur and provide a sufficient condition under which the rankings are guaranteed to agree. This result provides a practical heuristic for assessing whether offline evaluation is sufficient to reproduce the decision suggested by a live experiment. Third, we show, in a stylized linear model of interactions, that user adaptivity can increase variance in estimation. Finally, we discuss implications of user adaptivity for the interpretation of experiments and show that adaptivity creates a common mechanism leading to common evaluation challenges, such as reward hacking and novelty effects. Together, this work highlights the necessity of measuring both model performance as well as changes in the user behavior that the model induces.
| Subjects: | Methodology (stat.ME); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.17239 [stat.ME] |
| (or arXiv:2611.17239v1 [stat.ME] for this version) | |
| https://doi.org/10.48550/arXiv.2611.17239 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Hannah Li [view email]
[v1]
Thu, 12 Nov 2026 18:26:51 UTC (62 KB)
来源：arXiv cs.AI · arxiv.org